It depends on the villian... if the flat,if they fold, if they 4-bet, how big and often they are opening.I think its a good idea to start with a super strong range and then expand from there based on how villian plays.I actually start even tighter than that with 99+ and AJ+, I don't think I'd ever 3-bet KJ or KQ early on.I don't think a tight 3-betting range can be exploited, it may not be max EV but I don't see how villian can profit from you 3-betting strong hands. Also, it takes times for a villian to guess your 3-bet range. If you get AA on the 3rd hand and 3-bet, villian may flat 9J or shove A9, they don't have any idea what you have there.If you are going to expand your range (villian folds enough) its a waste to include hands that flop well like KJ, K10, QJ etc. Its better to 3-bet 83.I'm always amazed at how often people fold to my 4-bets, so my guess is that the a lot of the population are playing a much weaker range. And based on their frequencies my guess is a lot of people are actually only 3-betting their junk.Anyway thats me, but I'm a bit of a nit.

i think 3betting some lower suited conenectors is better than 3betting kj, kq hands unless villain is limping, then i like 3 betting those.

the deeper you are, the less of a mistake 3betting Suited connectors is going to be I think [ still villain dependent]. so the first two levels are an option to do this, but readless I tend to avoid it. Its a pain if villain likes to lead flops OOP, or has no fold button. vs stations: 3bet for value, they aren't folding often.vs TAGs : I guess a polarised 3bet strategy is a good idea against these guys.vs LAGs : 3bet for value, an unpolarised range is probably going to be good most of the time.