Combining some of the projects in this thread with that, there could be a completely new user management regime, and allow multiple regimes to exist on one forum. This could allow all sorts of filters on users. For example, there could be a list which bans people who regularly advertise, a list which bans people who sell certain dubious goods/services, a list of people who turn everything into a political argument, a list which bans people who talk very technically, etc. It would probably also be feasible to eliminate the posts of people meeting certain "hard" criteria, such as activity count. These can be tailored quite precisely to what a person wants and doesn't want to see, whereas management in forums now generally forces an admin to either allow everyone to see the persons post, or ban him, and nobody sees his posts.
I'd eventually like to eliminate most moderation and replace it with web of trust rating/ignore systems so you can choose whether you want to see trolling, profanity, etc. by modifying your trust lists. This is the sort of thing that pretty much
requires a forum rewrite, though: it's totally against SMF's nature.
Global ratings like Bitcointalk++ aren't usefully scalable. It's too difficult to prevent (and even
define) abuse as the number of users increases. Some sort of WoT is needed.
By incorporating a "plural moderation" script (this isn't being developed, afaik, but seems like an obvious extension from the ignore+ script and the OP script), this could allow multiple lists narrowly defining "abuse," or anything else someone may not want to read. Lists would be maintained by one or a few people who go through posts and mark certain posters with tags. With the ignore script, this could remove threads and posts from users with certain tags, based on what the user individually chooses to filter out. Posts may be able to be tagged individually, too, if someone feels up to it. For example, "dubious investments" could be a tag, and the moderators of that list could remove those threads from the forum without "the forum" needing to do any moderation on its part.
So, for example, let's say someone wanted to remove-from-view people who use referral links. Let's call the list "referral link spam." There would be 1-5 moderators of the list with mod credentials who could click a button next to a person's post (or through adding them manually), including brief reasoning for the inclusion in the list. This would tag them with "referral link spam" and remove their posts and threads from users' view who decide to exclude posts from people tagged with "referral link spam."
From a user perspective, you could choose whichever lists you want, based on both the criteria and moderators maintaining the list (there could be competing moderators for multiple lists removing posts meeting the same criteria if someone distrusts a particular group's judgment). So let's say the lists are:
*Spambots
*BFL shills
*anti-BFL shills
*libertaritards
*dubious investments
*Jews
A user could go into the extension or script settings and simply check off whichever groups of people they don't want - maybe BFL shills, Jews, and libertaritards. Anyone (or any post) with any of those tags would be removed from that user's view.
The most important benefit from this, I'd think, is that it gets around the dichotomy of abuser or non-abuser, and allows people to really choose what kind of experience they get out of BTCTalk. The service may or may not benefit from incorporating a charge system for access to a particular list, the fees of which might go toward paying the moderators of the list.
Re-reading this, it doesn't seem particularly clear. I can draw some mockups if it's too confusing (I'd prefer not to, though).