UTG: 47/29, RFI:36%, f3b 33, 4b range 2.5, 7k h
UTG: $67.84 (67.8 bb)
MP: $102.38 (102.4 bb)
CO: $85.85 (85.9 bb)
BTN: $100 (100 bb)
SB: $140.29 (140.3 bb)
Hero (BB): $110.58 (110.6 bb)
Preflop: Hero is BB with T
T
UTG raises to $3, MP folds, CO calls $3, 2 folds, Hero raises to $15, UTG raises to $39, CO folds, Hero raises to $63, UTG raises to $67.84 and is all-in, Hero calls $4.84
I would make my 3b sqz size smaller, 12 should suffice here.
Against the 4b it's tough. I wouldn't place too much emphasis on the 4b stat. Look at his 3b stat instead, since your sample size is low. VPIP and PFR seem to indicate that he's loose and aggro and will be jamming some slightly weaker pairs and AQ/AK.
If he really has a 4b range of 2.5 over a large sample (1k is very little), then you should probably consider folding. But given info provided I would probably stack off reluctantly.
Pot odds: 53/139 = 38%
Assuming 4B stat is unknown, i think this is a reasonable range
http://www.pokerstrategy.com
Equity Win Tie
MP2 58.73% 57.59% 1.14% { 99+, AQs+, AKo }
MP3 41.27% 40.13% 1.14% { TT }
If his 4b stat is 2.5, then maybe it might be too optimistic to add 99 and it might be a fold.
http://www.pokerstrategy.com
Equity Win Tie
MP2 64.16% 62.90% 1.26% { TT+, AQs+, AKo }
MP3 35.84% 34.58% 1.26% { TT }
He might also be the type of player to buy in for 100bb, but doesn't auto-topup. Then his strategy at different stack depths might be very different and it will be much more difficult to estimate his 4betting range.
Either way this decision is close and I'll probably stack off after making it 15. But if you make it smaller you probably can fold since you wouldn't be committed.
3b stat is around 15%, vs me much more, especially in late pos.
example he's small 3b vs me is 49(20) and big blind 25(29).
sqz i made it big, because Villain doesn't like to fold much to 3bets 
Ok i guess the 3b size is fine, especially if you are only squeezing value hands and you want to commit yourself anyway. balance isn't too impt vs the fish.
15% total 3b is insanely high, so you should definitely be ignoring the 4b stat for now, it's probably due to a small sample size.
pretty easy stackoff with this piece of information
