The Bloodhorse just ran an article claiming there is a "historical" advantage to to early speed horses in the Belmont. I think this notion of race specific data is nonsense. Ask any statistician. The sample size for a race run just once a year is way too small to be significant unless one goes way back which then gives you an unrepresentative sample of what's currently the case. There is nothing magical about one particular race that defies the laws of probabilities of all other races.
Yes the lack of early speed in this years Belmont gives speed horses an advantage but that is true for any race from $5,000 claimers to the Classics.