We could, as you say, analyze and figure out whether or not any given market "innovation" is good or bad, but the big problem is - we can't meaningfully act on such information. Coordination is superhard[0], and in the meantime we all make individual choices based on momentary value differentials - choices which the market will happily aggregate and amplify, whether we like it or not.

You're spot-on in splitting the problem into two in your other comment.

--

[0] - especially if you frown at one of its most powerful form in a large society - government regulations.