Although it is now not possible to translate SE into Chinese at this moment, I think it would still be good to offer some opinions from the perspective of Chinese so that the translation will be smooth once it is possible. Moreover, Chinese is not an Indo-European language while all the languages discussed in this thread are.
First of all, there is no space among Chinese words. But I do not see this is a problem, since you can add commas among the words, it will be as easy to remove all the spaces as well.
I think the most tricky part of the translation into Chinese is that many planet classes in Chinese should be ended with the type of the astronomical body - a planet, a dwarf planet, or a moon. For example, terra is translated as 类地行星/lèidì xíngxīng (lit. Earth-like planet) for planets, but 类地卫星/lèidì wèixīng (lit. Earth-like moon) for moons. Preferably, a terran dwarf planet should be translated as 类地矮行星/lèidì ǎixíngxīng (lit. Earth-like dwarf planet). While it may be not quite problematic to call a dwarf planet as a "planet," it will be rather awkward to call a moon "planet." It is understandable, but just awkward. As far as I know, in Japanese planet classes work in a similar way. Furthermore, It is not possible to call an "Earth-like moon" simply as an "Earth-like" because you need a noun to conclude the phrase just as in Russian as SpaceEngineer argued. To circumvent this issue, one may try to call all the terra - no matter planets, dwarf planets, or moons - as 星/xīng (lit. star), as a planet is a 行星/xíngxīng (lit. moving star), and a natural satellite is a 卫星/wèixīng (lit. defending star). But it is still unnatural and in this way, you lose the difference between a planet and a real star, which makes it tricky o distinguish a carbon star and a carbon planet. If we want to prevent such problem after the Chinese translation becomes possible, I think it is best to do something now. We could store a latent variable to indicate whether the astronomical body is a planet or a moon in the variable for the planet class, and conceal it for all the Indo-European languages where it is not necessary. Once the Chinese and Japanese translations become possible, the translator can immediately make use of this latent indicator.
Concerning the dimensions of classes to be shown, I prefer showing volatiles classes than surface classes from the perspective of a Chinese speaker, because it is much easier to translate. Surface classes are difficult to translate because words do not inflect and morphemes are granulated in Chinese, so a carbid carbonia will be a 碳碳行星/tàn tàn xíngxīng (lit. carbon carbon planet), which sounds awkward. Even worse, a carbid aquaria might even become 碳水质行星/tàn shuǐ xíngxīng （lit. carbon water-ish planet), which can be easily misunderstood as a "carbohydrate planet" in Chinese! (By the way, I feel aquaria is perhaps the most difficult one to translate, because in Chinese there is no distinction between the scientific Greek/Latin morphemes such as "aqua" and daily morphemes such as "water.") If one really wants to make the surface classes clear, the translation would inevitably become too wordy. For example, a a carbid carbonia can be translated as a 碳基岩碳行星/tàn jīyán tàn xíngxīng (lit. carbon bedrock carbon planet), which is not nice, either. But volatiles classes are much easier to translate.
Finally, I think there also have to be some adjustments of word orders in Chinese. But this should not be problematic since word order can already be changed.
A few examples for planet classes in Chinese in my mind:
Venus - 热类地火山行星 lit. hot Earth-like volcano planet
Earth - 显生水海温类地火山行星 lit. life-manifest water sea temperate Earth-like volcano planet
Mars - 二氧化碳冰川凉类地行星 lit. carbon dioxide glacier cool Earth-like planet
Io - 岩浆湖冷亚类地火山卫星 lit. magma lake cold sub- Earth-like volcano moon
Titan - 烃湖冻水质卫星 lit. hydrocarbon lake frozen water-ish moon
55 Cancri e - 潮汐锁定的岩浆海焦超碳行星 lit. tidally locked magma sea scorched super- carbon planet
Proxima b - 潮汐锁定的水冰川凉类地行星 lit. tidally locked water glacier cool Earth-like planet
Many of them still feel not right but the Chinese translation itself is something hypothetical in the first place, so we still have plenty of time o think about a nice translation. For now, I think it is just important to put the planet/dwarf planet/moon variable into the codes for planet classes for potential Chinese translation.