When there is a significant time discrepency we look at the horses to see which time makes more sense. We\'re not talking about a fifth or two here-- it\'s usually obvious who had it right. Also keep in mind Balletti is just one horse-- we use all of them, not just the winner.
The approach Beyer is taking with synthetic figures is interesting, and off the top of my head, wrong. If it were simply a question of some horses being robbed of good figures, when they jumped up on dirt far more of them would repeat those figures, rather than fall apart. It will make some first synth winners \"look\" like they make more sense, though.
I suspect that there is going to be a big difference in how this plays out with horses coming from California as opposed to other places with synthetics, and I think it will have more to do with testing in those areas than surface issues.