It is a bit too hard for me!
First Kotov Position in his book "Think like A Grandmaster"

rxg6 screams out to me to rip apart the pawn structure and utilize the b3 bishop. wait does nxf7 just win a pawn and rip everything apart as well

i attempted to get through the book a couple of years ago I remember it being exhausting maybe I should try it again

Nxf7 activates every piece on the board (except Bd2).
(opens the rook, undermines the e6 and g6 which "activates" 3 other white pieces).
Moves like that don't always work though, and you have to calculate.
And from what I know the book is all about candidate moves and calculating as if you were a machine.
Totally impractical, but it's a classic for... some reason

Since there are soooo many different lines you could calculate, I wonder what he says.
Because a lot of lines I'd probably go only 3-4 moves deep, and just get a feel for whether the attack is working or not.
And of course in a real game, you'd be distracted by lots of other options, like 1._xg6 or 1.Bxh6. Even if it's only briefly, it's still time off your clock.

And of course for every sac, there are multiple moves black could try.
So really the biggest factor in this position, I'd say, is not calculation. Of course calculation would be very important, but most important I'd say is experience. Experience with sacrificial attacks. If you can't get a feel for what is and isn't working, then you'll just aimlessly try to calculate 1000s of variations each 10 moves deep.

Very true!
With just a superficial glance at this position, I do not immediately see a solution. 1.Nxf7 does not seem as clear-cut as some people suggested.

Ng4 seems basically the same as Nxf7 (the threat is on h6) but Ng4 doesn't win a pawn.
I don't see how black defend Ng4. I'm just saying.
Anyway, back to your idea of black taking the rook
This is not easy