The patch for the menu settings within menus was discussed, but I think there was a lack of follow through from the person working on it, as often happens, and so it's somewhat faded away because there's no longer a force behind it trying to find what is necessary for it to reach acceptability.
The scroll wheel patch does not function with the new button action system, I believe, and until it's fixed to do so, can't really be included. Again, a case of needing someone to take it under their wing, and find out what the core developers ask of it.
The "The" sorting patch was not 'quietly', it was nixed as loudly as any patch that gets rejected. There's no reason why users can't simply name their files in a way that they're sorted effectively, and no reason to add the unnecessary addition to code size. Users have to remember that Rockbox is one piece of software capable of running on multiple targets and the limitations of other non-iPod platforms do have to be taken into account when it comes to basic functionality that will run on all players, such as track sorting. It was an option that was almost entirely a very mild convenience, and pretty much decided unnecessary.
What patches get accepted are discussed primarily in IRC, mainly because that's where the patch authors come to ask about them. Patches get accepted in the VAST majority of cases by the Author coming to the core developers, saying "What do I need to do to this patch for it to be accepted?". Now you may think that the core developers should be reading the patch tracker for patches, but really the core developers are working on their own stuff, and often only look at a patch when someone directly brings it to their attention, and will only commit a patch *after* it's cleaned up. The "apply, and then fix" methodology doesn't work for them. The developer's mailing list also works for discussing patches, but most authors seem to prefer discussion realtime.
So, basically what I'm saying is that for a patch to have a likelihood of inclusion, someone basically has to "sponsor" it by making sure the patch compiles cleanly against CVS, works, and meets the programming guidelines, as well as making sure to address the objections various people may have against it.