I thought I'd suggest a couple of modifications to the talent system that might be worth looking at. The main problem, as I see it, is that for a lot of talents, their effects are simply so minor that there's really no point in ever taking them. Moreover, most of the main benefits for talents are only really noticed in the early game. By the time you get to level 20 or level 30, new talents aren't really an important consideration. Here are a couple of proposals that I'd like to suggest:

1) Increase the potency of the weakest talents.

Of these, I think the ones in most desperate need of a boost are:
-All +Stat talents (charming, alert, tough, etc.): +1 Stat is a very minor effect, well within random variations for any character. I think if these were say, +3, they'd be pretty decent.
-Skilled lines: Very weak effect. Even taking all three of these is not really that valuable. I'd suggest the following: Change it so that taking skilled/very skilled/extremely skilled each add one dice to all skills. So a 1d3 goes to a 1d5 with skilled; a 1d5 to a 2d4, etc. This might be a pretty strong effect though. Maybe have only very/extremely give this effect might be a nice balance?
-All weapon affinities + brawler + ambidextrous: The problem is that many players change weapons frequently, and committing to any one style isn't always desirable. Also, +2 to-hit is still a pretty minor effect. I'd suggest either making the bonus bigger, or adding a damage bonus. Or both.
-Hardy line: Problem is that the characters who might want these talents (PCs with low To) can't take them. Maybe reduce To requirements slightly or increase HP gains.
-Silver Tongue line: Bigger effect would be nice here. Would also be nice if it affected items sold as well as items purchased.
-Pious line: 5% cheaper prayers is a very weak effect considering the quadratic growth rate of prayer costs. 10% each might be more reasonable. I don't know if this includes (pre)crowns or not, but if it doesn't, it definitely should.
-Mechanically inclined: This skill is rarely used anyway, so very few PCs would ever take this talent as it is. Not sure what to do with this one. Either scrap it or major overhaul.
-Scout/Stealthy: Increase the effect. There are skills that pretty easily cover these effects quite well as it is.
-Strong legs: Kicking is rarely used in combat anyway, even by monks, and this effect is not that strong.
-Healthy/charged: Both of these talents have very weak effects. The effect of charged is almost unmeasurably tiny for characters with max concentration, in fact.
-Good/great learner: After the bugfix on these talents, the effect is very weak. I'd suggest maybe 10% for good learner and 10% for great learner is not unreasonable.

2) Frontload talent choices toward the early game

At present, characters get talents every third level all the way up to 48, 16 in total, plus 1-7 starting talents. Since talents mostly are only effective in the early game, I might suggest the following: Have characters get a talent every other level between the levels 2 and 24, plus 1-7 starting talents. Characters would end up with only 12 talents from leveling, but they'd get them more at the point in the game where talents will have a significant effect, rather than having half of the game's talents being acquired so late in the game that their effects are no longer meaningful.

3) Create a small set of elite talents for higher level characters

An alternative to the above solution is to have a set of talents with a level requirement on them, but that are significantly more powerful than the regular ones. These talents could also have pre-reqs from the various talent trees, or high skill/stat values, encouraging players to choose talent lines from broader talent trees in order to get better elite talents. Or even combine this option with the one above: Characters levels 2-24 get regular talents every other level; characters 25-50 get one elite talent every 5 levels.