The main problem with Greek language and trancriptions is that we have a unique alphabet, not having adapted the latin characters (which is a good thing also as it makes as original keeping this miracle throughout the centuries). But for the developing teams that consists a quite challenging task to produce the captions, of course it can be done and I hope that in the near future Adobe will evaluate and maybe with the aid of AI we can predict it might be more easy to achieve. Europeans were beggings us to switch to a hybrid latin alphabet for decades so they can read our history and original texts more easily.
I do not think the Greek alphabet has anything to do with why it has not been implemented yet. I mean, if Whisper.ai can do it and has done it successfully, why can not Adobe? Perhaps there is an impression of insufficient usage-cases?
You have to realize the crowd/population target group of which you are referring, you just mentioned almost the 80% percent of the global population, the cost of developing the algorithm of such a unique and complicated language with countless idioms such as Greek language (spoken by only 10 million people) is something that a company has to consider before proceeding to a firm and reliable solution of such a translation. Did that answer your scepticism?