To confirm how widespread the practice of using Julian/Gregorian numbers for unconverted lunisolar dates is, I am undertaking a summary of academic sources on East Asian history.

I would like to propose the following addition to MOS:JG:

• Dates in premodern East Asian history are given in the appropriate lunisolar calendar. Lunisolar years are numbered by the Julian or Gregorian year in which they begin. When the Julian or Gregorian equivalent is certain, it may be included.^[a]

Potential adjustments:

• Replace "in which they begin" with "with which they overlap the most"
• Replace "numbered by" with "denoted by the number of" or similar

China used the lunisolar Chinese calendar for nearly its entire history – according to our article on the adoption of the Gregorian calendar, it was first replaced by the Gregorian calendar in 1912, though actual implementation took several decades. Years were denoted by the sexagenary cycle and/or era names.

The Chinese lunisolar year begins on Chinese New Year, which varies from year to year but always falls in late January or February;^[b] to my knowledge, this has been true since time immemorial. The lunisolar year and Julian/Gregorian are largely contiguous, with a maximum discrepancy of around two months. For this reason, most scholarly historical sources simply denote each lunisolar year using the Julian/Gregorian year with which the lunisolar year overlaps most, which is also the Julian/Gregorian year in which the lunisolar year began.^[c] This practice is so widespread and the discrepancy is so small that many sources don't even make note of it.

Though the modern Chinese calendar uses a very simple system for numbering individual days, as far as I can tell, the lunisolar calendar previously used exceedingly complicated systems to name individual days.^{[citation needed]} However, academics have compiled conversion tables for individual days from lunisolar to Julian or Gregorian dates, and some scholarly sources, including some (all?) volumes of the Cambridge History of China, have converted lunisolar dates to Julian or Gregorian dates using these tables.

However, much of Chinese historiography does not include specific dates for events. For example, most of the Shiji is written in a proseline-like style, giving the year in which events occurred. Since the lunisolar year doesn't correlate exactly with the Julian year, it is impossible to convert lunisolar years to the Julian calendar with certainty.^[d] Hence the need for clarification.

The Cambridge History of China was compiled over several decades, so certain practices have changed with time. Judgement calls such as romanization style are listed in the introduction area, though date conversion details has been omitted from several volumes. I have listed some volumes below in chronological order of publication, quoting their practices on date conversion where I could find them.

tldr; Broadly speaking, dates in the Cambridge History seem to have been converted to Julian or Gregorian dates wherever possible, using academically rigorous methods, especially in the more recent volumes.

• Volume 10 (1978): Seems to make no mention of converting dates. Romanization is discussed extensively, though.
• Volume 8 (1998), page xxi: "Dates have been converted to their Western equivalents in the Julian calendar until 1582 and the Gregorian calendar thereafter, following Keith Hazelton, A synchronic Chinese-Western daily calendar 1341-1661 AD, Ming Studies Research Series, No. 1 (Minneapolis: University of Minnesota Press, 1984). The reader should remember that when Chinese sources refer to a year alone, this year does not correspond exactly with its Western equivalent." Comment: The 1582 cutoff date matches MOS:JG. However, years alone remain lunisolar.
• Volume 5 (2009): "Note: Table 2 follows the conventional approach that treats the Chinese lunar year and the Western solar year as essentially coterminous, thus preserving the traditional count of years in each reign period. For a convenient reference see Chung-kuo li-shih nien-tai chien-piao (Peking, 1973), pp. 135-56. Reign periods were sometimes changed mid-year, to respond to or to influence significant events. For a translation of the Sung reign-period names, along with a conversion of their inaugural dates to the Western calendar by day, month, and year, see James M. Hargett, "A chronology of the reigns and reign-periods to the Song dynasty (960-1279)," Bulletin of Sung-Yüan Studies 19 (1987), pp. 26-34." Comment: Romanization is discussed briefly in the Preface, but calendars are relegated to this note under Table 2. Page number not given in the digital version.
• Volume 2 (2019), which uses Pinyin(!), does not mention dates beyond this: "Unlike prior volumes of the Cambridge history of China, this work will employ the BCE (before common era) and CE (common era) dating system, rather than BC (before Christ) and AD (anno domini) one. This has been done to be more in line with current academic conventions and to regard time in a more secular manner." (pages xvi–xvii)
  • However, it can be inferred that dates were converted using an academically rigorous method:
    • "A reign title, Datong,“The Great Unification,” was established to mark his accession on the first day of the lunar year, February 18, 535." (page 213)
    • "on the first day of the new lunar year (February 15, 557)" (page 224)
    • "on the jiazi 甲子day of the second lunar month (March 4, 581)" (page 234)
    • When the date is unclear, it is not converted, as evidenced by "a surprise attack led by Dou Tai early in the lunar year of 533" (page 191)

Explanatory Notes, page xi, The Calendar: "Dates are calculated according to the western calendar. Occasionally the word sui (meaning years since conception) is employed when the exact year of birth is unknown, or when someone's birthday is celebrated at an advanced age."

This means dates were converted somehow, and based on one spot-check this was done accurately, but sadly I do not see specified what source(s) they used for the calendar conversions.

Chinese History: A New Manual is "an encyclopedic and bibliographic guide to Sinology and Chinese history". It seems to be a good summary of many different sources on Chinese history and where to find them, though Seefooddiet identified a few inaccuracies and typos. It also gives an overview of various concepts in Chinese historiography.

The fourth edition, which I have access to in print, discusses calendars extensively in Chapter 39 (pp. 492–532). Most importantly: After the adoption of the Gregorian calendar most Chinese historical works have applied it retroactively to China's past. ... Finding the equivalent year in the Western calendar is relatively easy. Most dictionaries and encyclopedias such as HD or Cihai 辞海 contain tables of dynasties, names of emperors, accession dates, and era names and sexagenary graphs. But beware that many bilingual dictionaries (Mathews [1931] for example) do not indicate the difference between accession and the adoption of the nianhao 年號. Fortunately, dozens of accurate chronological tables (nianbiao 年表) are available (§39.13.1). §39.13 then lists the newest and best chronological tables and calendrical concordances. The chapters on specific periods of history also list tables specific to those periods, such as A Synchronic Chinese-Western Daily Calendar: 1341–1661 A.D. for the Ming. I think this is the single most important piece of evidence I have found so far: Lunisolar dates can and should be converted.

The fourth edition of Chinese History: A New Manual makes a very strong recommendation for the most comprehensive and accurate chronological table:

Yu Baolin 于宝林. 2010. Zhonghua lishi jinian zongbiao 中华历史纪年总表 (Comprehensive chronological tables of Chinese history). Shehui kexue wenxian.

This also covers all the adjacent states and minor regimes of China, which is very impressive. However, I don't have access to this supposedly legendary work – will look into that.

Chinese History: A New Manual's own editorial practices are listed on page xix, section SS.8 Calendar Dates: "When it is necessary to cite a Chinese lunisolar date, for example, 光绪二十三年庚申四月十五日 (the 15th day of the fourth month of the Gengshen [year in the sexagenary cycle] in the 23rd year of the Guangxu period) the form is as in the original (nianhao, year, ganzhi year, month, day) with Julian or (in this case, Gregorian) equivalent in parentheses: Guangxu 23 [gengshen]/4/15 (May 17, 1897). To avoid confusion, months with numbers are always lunisolar months and months in the Julian or Gregorian calendar are never indicated by number, but always by their familiar English names."

I am not as knowledgeable about Korean sources or history, but I will attempt a cursory summary regardless. For most of Korean history, Korea used the Korean calendar, which was very similar to the lunisolar calendar of China, possibly with minor variations. It is thus equally impossible to convert lunisolar dates to Julian/Gregorian dates when only a year is known.

• A New History of Korea^[1] (transl. 1984, original pub. before 1967): Makes no mention of date conversion, but gives years using Julian/Gregorian numbers and the suffix B.C. or A.D. This means years are most likely denoted as I have proposed.
• Korea: A History (2022): "Unless noted otherwise, all East Asian dates before the late nineteenth century follow the lunar calendar, which was the standard in Korea until the government adopted the Gregorian solar calendar on the seventeenth day of the eleventh moon of 1895, or New Year’s Day of 1896, according to the solar calendar." (page xiii)
  • The first use of a Western month looks to be page 222, for the Tonghak Uprising, which began in "February 1894".
  • Page 188 says that Korea adopted the Ming/Qing Chongzhen calendar in the seventeenth century and briefly describes its development and improvements.

According to Japanese calendar, Japan used a version of the Chinese lunisolar calendar from the 6th century until 1873. Unlike the rest of East Asia, the lunisolar calendar is not really used in Japan anymore, even for holidays.

• Should take a look at Cambridge History of Japan, Korea. Also look into Tibetan, Mongolian calendars later.