I remember watching one of ESL's videos (I think) of the National 24; he asked Wilko, who was a helper or marshal at that one, what he considers to be the key to doing well on a 24, and he said simply "concentration" (I may be paraphrasing). So a rider may be great at 10s, 25s, 50s and even hundreds, but they need to be able to concentrate on trying hard for 24 hours - not really an issue on shorter events, or at least not in the same way. Most people tend to find the Sunday morning the hardest part mentally - for me, it was the thought that you still have 8 hours of racing to go...the concentration wanders when you think of this and you slow down as a result.
I think it's very hard to project even a vaguely accurate prediction for performance on the 24 from shorter distance TTs. I guess you may be able to do so if you look at the times of someone who you know to have a similar 'performance curve' to you (e.g. you both do better on long distance than short). This year was my first 24, so through the year at club TTs, I was comparing myself to club mates who had done the 24 to try to see how I might fare - it proved pretty innacurate though. Admittedly, their fitness may have improved since their 24 or they may have been concentrating on short distance so that may have distorted the prediction.
FWIW, I think my most useful indicator was when I did the Heart of England 300 in April. I rode it on the same bike, with the same set up and at roughly the same intensity that I intended to ride the 24. It wasn't a very sociable ride, though I did ride with a few people on it, but doing that gave me a reasonable idea of where my fitness was and whether my target was reasonable.
100 mile TTs are of course useful as well, but of the two I entered, I packed on one (the shame) and the other went badly wrong due to dehydration so weren't particularly useful indicators of performance!