"Feat point" systems are pretty common homebrew projects in general, and this system in particular basically reinvents the exact feat point system that Sean K. Reynolds came up with back in the 3.0 days. Unfortunately, I think the feat point approach to fixing feats is fundamentally flawed, for three reasons:
1) Feat point systems in general are too granular.
2) Getting people to agree on feat values (or setting criteria in the first place) is hard, and most people have no idea what feats are really worth or what impact they have.
3) Feats are highly variable in value on a campaign-by-campaign and character-by-character basis.Granularity
When it comes to spells, it's pretty easy to figure out where something should go on a scale of 1 to 10—or, technically, 0 to 9—because (A) spells level are gained as the characters progress, so a given class gains access to all Xth level spells at the same level and they can thus be judged to be level-appropriate or not for Xth level, and (B) spells range in power from "make your porridge taste slightly less bland" for cantrips to "ask God a favor and he has to do it for you" for 9th-level spells.
Neither of these apply to feats. Regarding point A, a feat worth, say, 8 points might first be able to be gained at 1st level if it has no prereqs or 20th level if it has a BAB +20 prereq or the like, so there's no common basis for comparison across levels and the difference between a 6-, 7-, and 8-point feat is huge based on level gained, number of prereqs, etc. Regarding point B, feats have a much lower average power ceiling; things like Epic Spellcasting and Epic Leadership are certainly incredibly powerful and game-changing, but there's a huge
drop in power from there to things like regular Leadership which is on par with a mid level spell or two down to even powerful fighter-centric feats that work out to low level spells at best, so feats are weighted heavily towards the lower end of the scale.Criteria and Values
I don't mean to sound overly critical, but the examples you list illustrate both issues pretty clearly:
- Tier 3 "unlocks new avenues", but four of the five examples are purely numerical boosts (and small ones at that) that just reinforce what you're already doing.
- Tier 4 "refines specialization" and gives "moderate benefits", but this tier include both Toughness and Improved Initiative; the former grants a negligible benefit that one won't even notice past level 1 or maybe 2 and the latter grants a universally useful and noticeably-large bonus to a stat that everyone wants and is hard to boost, so it's hardly a "moderate" benefit for either one, and neither one is about specializing at all.
- Tier 5 "impacts your ability to succeed" (which literally every feat does) or "opens a new option of impressive power" (which none of the listed examples do, making it hard to judge what "impressive" is supposed to mean here).
- Tier 7 "provides major quality of life improvement or new options", neither of which the listed example does (DR 2, even 2/—, is fairly trivial past the low levels and doesn't grant any new capabilities).
- Tier 8 "provides game changing modification to character or noticeable party change", and yes, Vow of Poverty does do that...but only in the sense that a character who takes it is pretty severely restricted without items and is noticeably weaker.
Those are five different tiers that all claim to do basically the same thing—grant generic numerical boosts or grant new options—yet there's nothing to differentiate them aside from some vague adjectives and none of the listed examples clearly fit into those categories at all.Variability
To return to Vow of Poverty for a moment, it's a terrible feat. VoP doesn't provide anything comparable to the benefit of items in terms of either power or options, characters with the feat rarely meet the sort of "must be this tall to play" level-appropriate benchmarks, many classes are weakened at best or crippled at worst without items even if the raw numerical bonuses from VoP are level-appropriate, and so forth. By the listed criteria, it's a Tier 1 feat: you take it if you were going to roleplay a peasant without two coppers to rub together anyway and don't want to be completely useless, otherwise don't bother.
you're playing in a game where you loot is very limited (survival horror, fighting lots of monsters without gear) or overly restricted (DM is stingy with loot for you or the whole party, a module with random loot tables where everything's subpar) anyway and you're a monk or druid or other class that can survive without the limited gear you get, in which case it's a no-brainer pure power-up that everyone wants to take and 1 point is far too low a cost for the benefits.
The reverse goes for Leadership (usually very strong, occasionally useless instead of usually useless, occasionally very strong), but it's the same general idea: there's a big difference between taking Leadership in a campaign where armies are useful, the DM lets you build your cohort and choose your follower, you're doing lots of roleplaying and politicking where contacts are useful, and so forth versus taking it in a campaign where you're dungeon-delving with no room for many NPCs, the DM randomizes your cohort and followers or builds them for flavor, and so on. Item creation feats are somewhere in the middle: usually of moderate power, amazing if you have lots of downtime and/or can use cost-reducers, useless if you're running from one adventure to another and/or get custom-tailored gear already.
On a character basis, some feats have certain synergy with class abilities (metamagic feats are better for full casters than for half-casters because they have the slots to use them, Martial Study is better if you already have an initiator level from your class and so can get higher-level maneuvers at a given character level and recover them in combat), some have synergy with other feats (Psionic Body with lots of other psionic feats, Arcane Thesis with lots of metamagic), and some are just very niche (Lightning Mace is a weak-to-mediocre feat on its own, but with critical enhancers, aptitude
abuse, and the like it gets really ridiculous).
The only way to fairly price these sorts of feats using such granular and variable criteria as feat points so that characters don't either snap up tons of especially-good-for-them feats for 1 or 2 points apiece or have to spend 7 or 8 points for marginal benefit would be to give the feats differing point values for different characters based on their class, race, build, other feats, etc., and that's a rabbit hole that no one is ever going to go down successfully.
As the very first line of your post says, feats are too diverse to categorize them overly strictly. Using the same universal "1 feat slot" currency is a poor way to organize them, but trying to shoehorn feats into an at once overly-rigid and overly-broad system like feat points doesn't improve on the matter at all, it just goes from too few incoherent categories to too many incoherent categories.