Here is an idea I can get behind. There are three basic parameters for each top level thread. Age, upvotes, and number of replies. Using a user-defined, weighted average to decide what is seen is a slick idea. I can envision the default being set to X, and then auto-updating with a simple average of all the users' preferences. Probably only dedicated users will likely take the time to use this feature, but that could be a good thing, because it will mean that only users who really care are contributing the the crowd-sourced optimum.Hubski could easily put its parameters on sliders and throw them up on the user page... or even on the comment page itself. Let users fuck with it to get what they wanted.
I think "probably" is likely to change as the community changes. It was an off-the-cuff suggestion to illustrate how the tricky, human-centric job of "making the site more usable" could potentially be automated. I'm with Vinod Khosla, as I've said before - "If it doesn't scale, it doesn't matter." I think any time you want to do something, if you say "would I want to do this for 100x as many things as I'm doing now?" you're likely to figure out the easiest way to do something. Either that, or decide that there's an upper limit to the size of project you want to support. That, I believe, is the tragedy of Reddit - when Alexis and Steve built Reddit, they weren't thinking in terms of millions. They were thinking in terms of "if we can get 20,000 people to sign up we can sell this thing." by the time Conde Nast had it and started thinking in the millions, Alexis and Steve were thinking in terms of "eighteen more months and we're vested and out of here." Maybe Hubski tops out at 1500 people. I hope not, though. And I know the only way to ensure that it doesn't is to think in terms of 15,000, 150,000 and 1,500,000.