Regarding mining centralization and problems that it brings, BBR should keep current trend and keep smaller miners as the hashrate increases. A step toward that is, after the pool functionality and implementation converge sufficiently, to give simple tutorial on setting GNU/L. pool and mining on a local network. Ppl will figure out what to do with it and will form associations, ideally a large number of guilds of different sizes and sociological types, to reduce variance. BTC pool mining forming model was not the perfect one, so approaches to just postpone centralization should not be discarded.

Regarding the price in such a constellation, if guilds would have orthogonal interests (offc that all have to pay electricity, but all have different situation in e.g. electricity price, not to mention politics), i suppose that if guild sizes and types (their particular interests, communication channels with language barieres, etc.) is large and ortogonal "enough" - it makes price only more fair (and stable) as it would "average-out" some of FUDs & hypes (negative & "positive" propaganda). Even in non-ideal case, to a certan level it would probably mitigate or filter out some "higher frequencies" in socio-driven cycles.