I think you completely misunderstand what they're trying to do. What you read as a proliferation of crap is really trying to find a 'big hit' in lots of small niches. Some niche hits will become crossover success stories, and draw people into new genres, etc. But others will quietly be classics in their small niche, often for being the first serious effort.

Of course, Sturgeon's Law is still in effect, so 90% of everything will be crap. But producing non-crap for smaller audiences builds loyalty, which is worth something in the subscriber model.

Or at least that's my read of their strategy. There's a real discovery problem that comes with it that I don't think they've solved. But overall it's an interesting approach.