I don't think Kasparov retired because his rating was dropping and he wanted to quit for reasons of vanity. I think he retired because (a) he has strong national-political beliefs right now, and thinks he can do something important with them (b) chess stopped being a challenge for him - there was no World Championship to win, and the thrill of winning the three super-GM tournaments he entered each year had probably gone.
This thread raises an interesting question though in my mind, which is how to measure 'greatness'. I think there is a lot of talk about this that doesn't mean very much. My thinking about it, which also may not mean very much, goes as follows. Pre the 20th Century, chess wasn't organised or known enough to measure the outstanding players against 'all time'. Philidor and Morphy just didn't have sufficient rivals to realistically discuss them. They were greats, but too far back. It's like discussing music before orchestras. After that, chess achievement becomes increasingly measurable, due to its literature, organisation (eg World Championship process, prestigious international tournaments) and relative transparency. What I mean by the latter, is that especially by or at least after the 1920s, there was not much more to be known conceptually than is known now. Hypermodern, classical and dynamical principles were all evident, if only starting to enter play and chess-writing. This transparency reached its clearest expression in Fischer, who could play in nearly any style - his style is enigmatic in the sense that his best games showed many dimensions of chess - attacking brilliancies, opening sophistication and preparation, diversity systems and position-types, different winning attempts at all stages, etc.
So I think in assessing the all time greats, we can only begin to get objective at the earliest with Steinitz, although more probably with Lasker - who as Kramnik says, was the first player to employ dynamics, albeit without the opening repertoire to really show this. (Lasker's dynamic play often came from cramped 1. e4 e5 games, and his openings don't really stand up now, whereas today the KID, Benoni and Sicilian which were developed really after his time would be the natural place to find this style.)
Asides aside, my conclusion is that we can realistically discuss greatness only in the 20th & 21st Century. How? My way of thinking about this is in two ways. (1) Volume of achievements. (2) Peaks of achievements. If you see the careers of the greats as something akin to mountain ranges, it's obvious that some mountain ranges stretch farther but not to world record heights, whereas others have huge spikes but reach less far.
On this basis, my way of positioning great chess players according to (1) and (2) would then go like this. (1) The volume of greatest achievements are to be found in the achievements of Lasker, Karpov and Kasparov, as any one looking over the stretch of their rules and ratings will see. (2) The greatest peaks of chess achievement were those of Fischer, Karpov and Kasparov. Fischer - the winning run in the Candidates matches, staggering. Karpov - Linares 1994. Incredible. Kasparov - well, the list here is just huge.
And that's where I stop, and get on with my terrible games instead.
PS. As yet, Topalov doesn't register at such heights. Maybe one day. Maybe one day for Kramnik too. Or maybe as chess organisation disappears and loses credibility, such heights will only be left for Hydra et al to automatically and uncaringly scale.