So let's see if these have been fixed. In order, by poster, they are as follows.
The peaks of TO: Yup
Caster Supremacy: Yup
Correct Tanking: Yup
Learning Curve: is a thing ... see 1)
Balance range: Yup
Limited combat feats: Yup
Few rules for social interaction: is that a bad thing? Do we really want more niche rules?
Magic Item dependency similarities: see 2)
Class ability value vs ECL cost: Yup
Combat flow: Yup. This would require turn phases (doable via the free/immediate/swift/move/standard hierarchy). Would anyone play this way, though?
Bad writing: see 3)
1) In other games (like magic the gathering), this is a good thing. I think it's acceptable in an edition. The only real fix is to start banning things that are too weak, which always steps on someones toes because they like the flavor.
2) Wait so we like optimization, but not when it converges? That's like saying you don't like x video game because the play through tends to follow the same route when done optimally. Well, yeah. If you don't like it, don't play optimally. I think it's asking too much of a system to be that balanced and diverse. The only real option is to soften up the worse choices, since nerfing powerful but balanced ones causes ire.
3) Guilty as charged. The only way to fix it would be to privatise (give creators full power over) the content. When "thought bottle" has only 1 guy's name on it, then we know who to ask to fix it. In the meantime we have this guy.
Btw the shamus young link shows that the DM lacks system mastery. Legolas could in fact do those things in D&D. He should too, because D&D is not a video game. If anything D&D is more suited to being shadows of the collosus than shadows of the collosus is! Video games have inherent limitations about what you can climb or kill. D&D doesn't.