By the time we can realistically expect colonization (around the early 19th century) , written communications in Chinese are too standardized and socio-economically entrenched to diverge it short of mass reeducation camps
Not quite; even during the 18th and 19th centuries, Written Chinese was changing and was not that entirely entrenched.
Wényán was becoming more and more eclipsed by
báihuà, despite the best efforts of the Imperial authorities, and even then basing it entirely on Mandarin was not that certain - not to mention the shite-ton of variant characters and alternative writing styles which distorted the shapes of the characters, the fact that even in Mandarin there are morphemes which do not have adequate representation in characters (and even more so the case during the early 19th century), and illiteracy was very high. It was even possible to not use characters entirely, as use of Arabic-based scripts among Chinese Muslims or the use of Nushu demonstrated within China itself. On top of that, dissatisfaction with the Confucian-style educational system which undergirded the characters was growing - which could be easily used by the colonizers. So Written Chinese is not as monolithic as it looks at first glance.
(and the magic of written Chinese is that it is pretty interchangeable no matter the dialect, since it isn't a phonetically based written language).
Wrong. Even in its modern Vernacular form, Written Chinese is not that interchangeable (and no, Chinese is not made up of "dialects" - that's a false interpretation of the regional varieties and an incorrect translation of
fāngyán which misses that in objective analyses they are actually closer to languages, so the preferred terminology now is to reserve "dialect" only for what used to be traditionally referred to as sub-dialects and use compromises like regionalects or topolects to cover the upper tiers) and is predicated only on knowledge of Classical Chinese before the Xinhai Revolution and on Modern Standard Mandarin (often inaccurately called "Standard Chinese") after the Xinhai Revolution. That places non-Mandarin speakers at a disadvantage and makes it equivalent to learning a foreign language, creating a diglossia sort of effect. What would be needed here would be to disassociate Written Chinese from its Northern bias, which can be done because the characters themselves, in actuality,
are grounded on a phonetic basis. To be fair, the phonetic basis is on Old Chinese, and thus the phonemic-grapheme correspondence is as problematic as is, say, English, but it's still there. If any interchangeability exists, it's on the basis of the phonetic core of the characters, which once formed an unsystemized ancient syllabary that could have formed the basis for an even more standardized written language than now if the number of characters was kept very low at the beginning. All written languages are meant to represent speech of some kind, and it does show by trying to find some way of representing how to pronounce the characters, since speech is always prior to writing and speech dictates how the written language eventually changes. In this way, that Chinese characters are to syllabaries what English orthography is for alphabetic languages does not take away that at its heart, there is a phonetic core which makes the characters no different from other written languages - indeed, when new characters have been historically formed, or when old characters (standard or variant) are reappropriated for representing foreign loanwords, from Sanskrit on down to European languages and Japanese, it is almost invariably on the basis of phonetics (for example, how else are there two different Modern Standard Mandarin renderings for Barack Obama's surname, with
Àobāmǎ preferred in China while
Ōubāmǎ is preferred in Taiwan?).
So Written Chinese, then, is not that special, and under colonization could be used, on its own and/or alongside Romanization, as a family of scripts no different from the Brahmi-based ones once the phonetic bases undergirding them change. For example, a Standard Written Cantonese could emerge on the basis of regional calligraphic styles and variant characters and promoted as the
bahcwuá for that part of southern China, and it would look nothing like characters and/or Romanization elsewhere (Romanization would be preferred by the colonizers, but alternative character-based orthographies would not be a problem as long as they're not based on the Imperial standards). Or it could be using a different phonetic-based script akin to bopomofo.