It was mentioned in an earlier post that sprinters will slow exponentially as they get beyond their distance preferences. This is true.
What is also true, however-- though, as another poster noted, facts like these might not be popular truths given the long-distance slant of this board-- is that distance runners' performances fall off sharply when they reduce to 800/1500/mile and, especially, sprints. If you have the slow-twitch capacity to be the WR holder at 10k, there's no way that you will have the fast-twitch power to be a top half-miler, never mind trying 600m.
Bekele has never broken 3:32 for 1500m. With specific training, I think it's fair to say that he could get down to 3:30, but that's about it.
Bekele's 24.5 final 200m was at the end of a race, yes, but it was also accomplished with a flying start, not reacting to a gun from a dead standstill. And let's face it, El Guerrouj went past him without any real trouble in that 2004 race. El G was throwing his arms in the air in celebration before the finish line. The assertion that it was only one stride length difference at the finish is false; it was a full two strides. There's no comparison in terms of basic speed, and El G was a 47 400m runner (there's a reason why his secondary event was the 5000 and not the 800).
Renato Canova, who knows a thing or two about running, once posted that he didn't think that Bekele could break 50. I'm a little more positive about his speed, but I think 49 is about his limit. 48-high, tops.
Bekele is a great runner; perhaps the most dominant long-distance runner to-date. But he is a distance runner, not a middle-distance runner, and at 600m-- which is the no-man's land between middle-distance and sprint-- he would totally be a fish out of water. Bolt would not relish the added distance but he would have to utterly collapse to lose to Bekele at 600m.