Phonetics isn't my strength (my ear for these things isn't very good), so I don't have any strong personal convictions on this. And of course there's generally a huge gap between the theory and actually being able to apply it in one's own pronunciation.
Some of the details about vowel quality etc. are still hotly debated in academic circles. If you're interested in this, the standard book on the topic is W. Sidney Allen's Vox Graeca
which gives the evidence for and against various interpretations. It's a bit technical if you don't have any background in phonetics, but I assume from your post you have some basic phonetic knowledge. (He also has a book on Accent and Rhythm
in Greek, but I'm not sure if it's been superceded by more recent scholarship)
Pronunciation is also something where you are going to get a different answer depending on what time period you're talking about (there's more than a few centuries separating Homer from Plato).
There's still a lot of disagreement about the grave accent in particular. I think it's often considered to be a neutral tone rather than a falling tone because of manuscript traditions which used it for all unaccented vowels, but again, I'm not au fait with all the details, just the general outlines of the debates.
If you read German there's an interesting article by Danek and Hagel that suggests the pitch change is not really restricted to the accented syllable, but spread across the word/phrase and that the syllable represents a pitch peak. This made a lot of intuitive sense to me but I don't know how it compares to other current theories about the pitch accent.
I have vague recollections of reading somewhere that Japanese may actually be a better comparison than Chinese for understanding ancient Greek pronunciation, although I can't find the source at the moment.
Anyway, the difference I see between Chinese and ancient Greek is that Chinese doesn't have a pitch accent -- it's a tonal language, meaning that tone is used systematically to distinguish lexical meaning. Greek doesn't do this. So while a Chinese speaker might be more comfortable producing pitch variations at will, they're doing so in a very different context than would have been the case in ancient Greek.
I've been collecting resources here if you want to do more reading (I need to update some of the links, but the articles at the top should all still be fine):http://spiphanies.blogspot.com/2009/03/ ... greek.html