Influence of chess computers on thinking process

By solving a combination from my favourite combinations book nowadays (Tactical targets in chess by Istvan Pongo) I found very interesting detail in solution of next position.

Black to move

 Of course combination is not hard and Black wins by:
1...Bxf2+ ...which is typical geometry motive. There are four possible moves, but three of them are real candidates.

a.) 2.Kxf2
b.) 2.Qxf2
c.) 2.Kh1
d.) 2.Kf1

I will show short lines in every variation with special focus on variation D.


a.) 2.Kxf2, Qf4+ 3.Kg1, Rxc1 .....and Black has a two pawns more.
b.) 2.Qxf2, Qxc1 3.Rxc1, Rxc1+ 4.Qf1, Txf1+ 5.Kxf1 ...and again Black has a two pawns more.
c.) 2. Kh1, Qxc1 ...which shows that 2.Kh1 is not a real candidate move (it is only a legal move).
d.) 2.Kf1! This move is not analysed in the solution and I think that this is white´s best option. It is not so hard to spot this move but it is somehow visually illogical. Such moves are very easy to overlook during calculations but analysing games with a computers made humans used to think about moves like that. It is not that a decent player can not find a move like this. It is more about when calculating long and complicated lines during a games we usually makes some inaccurate moves and there are a reasons behind these mistakes. Like I said, a move 2.Kf1 is visually illogical, and I am sure that Pongo would give it in the solution ,if he actually spot it. There is only one good move for Black now but it is good enough for comfortable win.
2...Bc5! Idea is Qf4+ so....3.g3, f4 ...with a threat of Bh3 mate 4.g4, e3 and White should resign.