I actually audibly sighed because I was really, really hoping something would show up.
It might not be a measure you agree with, but it’s even the same one you mock Sentry for in this post. An average. Wow.
To answer the question, however, all three statistics are potential indicators of issues and can indicate different issues and have multiple possible solutions that do not cause other fields to, themselves, be weird.
To correct Terran whole-population average MMR being low, you’ll basically have to significantly buff the race when not micro’d properly, which will very likely cause them to be absurd in Masters and up.
This is because, to you, Terran is being bogged with extra weights and barely functions, it’s the weakest race, right?
That’s not right and you know it, directly from your list of three statistics in this post.
“In the top 0.1% of players, Protoss is the worst.”
“Averaged across the entire population, Terran has the lowest MMR.”
Both of these do not necessarily indicate a problem. But I think that it’s obvious that the former would be more likely indicative of an issue - intrinsically, because you’re looking at the game’s best players.
But that doesn’t mean there are or aren’t issues.
But what is important to realize, is that whole population averages are frequently wrong and you need to use some other analysis in order to make anything useful when you heavily oversimplify data by boiling it down to a single number.
I simply don’t know how to communicate this in language you understand.