Of the original options, I prefer #1 and #5. MAAAAYBE #2, but I think it, and #3/#4, would create a huge step in power between mechs with >= X damage and mechs < X damage.
The problem with the straight-damage method, in addition to being a buff only to mechs with high damage is this: either you have to set the threshold low so that a reasonable percentage of units could take advantage of it and have crits getting thrown out like armor didn't exist, or you push the damage threshold up, and reward mechs that are already blessed with high damage values! Kind of a catch 22 unless you could find the perfect damage threshold...but I don't think there is one. Also, if you do pick a low threshold, now it sucks to have high armor/structure because you have a serious risk of becoming useless well before you blow up.
I like #1 for a couple of reasons. First, I consider the Alpha Strike to-hit roll to be sort of the TW to-hit and location rolls smooshed together into one. In TW, there is a 'magic' 1/36 chance of getting a TAC on a mech. It happens to be 2, but 12 has the same probability, so to me multi-purposing the to-hit check to also be a TAC check is fine. I really don't care that it makes landing a low-probability shot more damaging. I admit there is nothing in TW that gives you an extra reward for landing a really hard shot, but I don't think it would actually change much. We are talking about an extra crit, on average, once every 36 to-hit rolls. That just isn't a whole lot of extra crits, and it is super-low effort in terms of adding rules and complexity to a game that is trying to stay light and tight. Finally, there is no penalty for trying for low-probability shots anyway in AS because we don't track heat for regular shots. I ALREADY will try any shot that is mathematically possible, even without the possibility for a TAC, so adding this rule isn't going to change the number of low-probability shots people try for.
#5, while I don't think it addresses the core problem, will make crits stick more often which is good. Unfortunately, I don't think it will change much on its own because it doesn't change when the crits happen, and therefore how long you have to deal with them. If we change other stuff, we may not need this one.
That said, I am in favor of having crits show up more often earlier in a mechs lifecycle. I also agree that, as-is, crits really don't matter much. In my experience, most mechs are about one turn away from dying when they finally get a crit. non-XL assaults might get two or three turns. As someone else said, in 3025 crits can matter, mostly on assaults and some heavies. Once XL's show up, structure drops like a rock and crits don't happen early enough in a mechs life to make much difference.
I have more to say, but I need to get my thoughts in order...and eat lunch. the short version is, I think we should, conceptually, have a goal for how many crits we would like a mech to accrue before it dies based on it's total armor/structure. If we have a goal, we can better evaluate how different options would get us closer to that goal...or how they wouldn't.