xombie: Hardly. It's a masterpiece. Take a look at the moves between 20 and 25. White has a good queenside attack going, but black's counterplay (as always in these positions) comes from playing f4 and opening up the e file.
Petrosian's defense is ingenious and instructive.
20. Qb4 kicks the black queen, and black puts it on the e file. Perhaps that was a mistake. Maybe Qf6 right away. Note also the diagonal defense of e1. 21. Re1(!, !!) indirectly prevents f4. It is very clear that white had anticipated Re8 ... probably several moves ago.
This is a brilliant shot. It prepares the combination that we shall see in a few moves, to whit, Rb7, which is key. And so, the queen gets out of the way for the rook. In many ways it is also a waiting move, because white wants black to commit to doubling on the e file, at which point white cannot take on f4 (after f5-f4) if he plays Rb7. But if black does nothing, he just sits. Furthermore, now the Q defends f3 as well. This had escaped me immediately (after all, what is the point of defending f3 at the moment). But look forward to move 28 and the importance of this point becomes manifest. After f5-f4, black not only threatens to open up the e file, but he also threatens f4-f3! This must have been critical in Petrosian's evaluation - the consequences of refusing to take on f4 and instead playing e3-e4.
22 ... Re7 (obvious and natural)
We can see that now white cannot take on f4 after the pawn push. Therefore, he is quite committed to the Rb7 plan, should Re8 occur.
I think this is another waiting move. It is nonetheless very useful, as it forces black to show his hand. It also removes the B from future threats on the f file, and guards h3.
23 ... Rfe8
And black now commits.
Obvious now, given our line of reasoning. But note that white only plays this after black commits. f4 has been prevented, as is shown by the continuation.
24 ... Qf6
I wonder why black did not play this right away at first. There might have been a reason for this ... one does not know.
This is both a pawn-eating move and a waiting move at the same time. Then again, black has nothing if not f5-f4.
25 ... f4
The much anticipated breakthrough, which is refuted thus:
My !'s and !!'s are all perhaps superfluous since we have already outlined the reasoning behind these. Be that as it may, we can see that the breakthrough fails tactically.
26 ... dxe4 27. Rxe4 Rxe4 28. Bxe4
We can now see that 28 ... Rxe4 fails to 29. Rxd7 with another set of pieces being traded, and a hopelessly lost game ensuing for black. Also, now that we are here, note that f4-f3 does not work because of the Qc3 (that innocuous 'waiting' move from long ago)! If Qc3 wasn't there, black might have been able to play, post B trade, f3, Qe6 (... Qe6 Qb3 ... Re1 etc). Finally (in the game), the Q moves to e1 to defend g3. However, I cannot tell if Petrosian had seen this far. It is certainly not too difficult, but enormously imaginative.