It doesnt matter for current hand whatever has happened on previous hands card distribution wise. Maybe easier to think of game like roulette. Next spin has same probability of it being red whatever there has been 2, 5 or 20 times red in row in previous spins. But to say that from this point on next 10 spins will also be red is very unlikely.
For ranges they can become wider if player is mentally thinking of short handed situation requiring wider ranges. But why would it matter if you are BU in 3 handed or 6 handed table? You still have only blinds to play against. Outside of that rake side which might change something gto wise.
Just to check hand like KK53ss EV in trainer if facing 4bet we will get this:

So folding we will lose those 11.5bb we already invested. Calling would lose around 5bb more while shoving would be 12bb more lost compared to folding. Those are enormous differences in single hand. E.g. if these EV assumptions would be correct on average then our EV for 5bet shoving KK53ss would be -1200bb/100 hands.
In general I would say average player has more AA than trainer has which would be detrimental for our EV. We could ask how much of BU 4bet range is AA then in trainer? In plo50 sim BU is 4betting 8.9% versus SB 3bet. AA will always 4bet and AA makes 5.3% of BU opening range. So proportionally its around 60% from 4bet range.
KQT$ds is not really combo that should 4bet so this could be indicator of villain over 4betting. But it could also be his way of overvaluing this kind of combo and he might lack lower ds rundowns that are actually 4betting gto wise.