There's one big factor you haven't taken into account and that's the changing demographics of the Pazz & Jop Voter. There used to be far fewer critics voting in the poll and many of them represented more mainstream media like daily newspapers. Most of those jobs don't exist anymore and I would guess that as P&J adds new voters, most if not all of them come from the blogosphere - younger, hipper, and far less involved with Billboard positions than the aging, retiring, or unemployed dead-tree media critics they replace.
A very fair and apt point--thank you.
I still say, though, that the dropoff in sales from the three indie-centric winners from 2007–09 is too steep to be explained by a changing voter panel alone. It's always tempting to point to a big macro trend like that and say, "It's mostly this"; but even as rapidly as the Internet has Changed Everything in the last decade, if you can point to a counterfactual in the last 3–5 years, that argument is a bit less persuasive.
I had a similar (pleasant) debate with some American Idol fans who responded to my October column about Scotty McCreery by explaining away his decent-but-modest sales as the product of the industry-wide downturn in album sales. Yes, I said, except there were a couple of Idols who sold significantly better just three years earlier – album sales since 2008 are down, but they're not that far down. (They're down far more steeply since 2001.)
OTOH, to give your argument its due, I suppose it could be argued that the shift in dead-tree criticism in just the last 2–3 years is staggering, an even more precipitous shift than the steady but slower downturn in music sales over the last decade. So you definitely have a point there. My only point is, let's not overstate the effect of one macro trend.