a) This is a spot I used to raise, but I now think completing is usually better. If there are a lot of short stacks (50-80 bb), then I don't mind raising with the hope of isolating one or two of the short stacks. The reason for this is that many short stacks will call a wide range preflop and make a commitment decision on the flop, so you have a chance of winning the pot with a small cbet when they miss the flop and give up and can also win their stack when you flop TPGK and they decide they are going to continue with their hand. Basically, short stacks often play terrible and you need to be aware when you have someone who is looking to "double up or go home". The other time I would raise is in a GREAT game where literally every hand is seeing 7 or 8 to the flop and every raise is being called by everybody. In that case, I'd raise to $20 or so with the intention of trying to get all-in on the turn if I flop TPGK. In general, KQo plays poorly in a multiway pot against competent hand ranges, but against random hands that include trash like 64o, K7o, J2s, etc. with an SPR that allows you to get the money in before the river, the situation becomes very profitable. But my default gameplan is to complete.
b) I'd complete here too in general. As other have explained, the dynamics of this hand plays better in a limped pot than in a raised pot, especially since we are out of position.
c) I'd actually strongly consider making a small pot sweetener raise, to about $7. I don't think you'll get many folds and that's not the point. You're giving up a bit of equity, but you bloat the pot so it becomes easy to get stacks in if you do flop your set. If the stacks are deeper, I'd consider raising even more (for the same reasoning).
d) I think this is an "it depends" situation. Is the button sticky (calls lots of raises preflop) or will a raise often buy the button? Will a large 8x-10x raise likely result in a shorthanded pot / taking down the pot, or is it likely to bring about a huge bloated pot? I think overlimping, making a smallish value raise to $10 or making a bigger raise to $18 or so could be right based on table dynamics.
e) Similar reasoning as with d.
f) I'd again make a small pot sweetener raise. We give up a bit less equity because of our position (we may get checked to and have a chance to take a free card on the flop) and we again give ourselves a chance to play for stacks if we floparoo a setamagoo.
g) I think raising generally fares poorly here. I think you can limp here if the game is really good (specifically, very little raising preflop by anyone but you, and people playing straightforward and loose postflop), but folding is probably going to be the best play the majority of the time.
h) I'd still make a pot sweetening raise, though this one is the most likely to backfire because someone 3bets. But in actuality, live players just don't 3bet without premiums, even against transparently small raises. So if someone has TT on the button, if everybody limps in to him, he might raise to $15 or $20 to try to win the pot there with the "best hand". But if you make a pot sweetening raise to $6, that same player will often just call with everybody else. So your super exploitable raise actually creates a much better situation for you when this happens. Of course, sometimes someone behind you wakes up with AA or KK and they reraise big, but that doesn't happen too often.
i) Similar to g, but this is not a very good hand. You should be looking hard for reasons to fold. Only in a super soft game (where I'm limping almost anything suited or connected in late position because the game is that good) would I be looking to get in here.