Some thoughts:
First, I'm certainly still interested in data if folks want to give it a go.
Second and perhaps more importantly, though, I think you all have collectively convinced me not to mess with the "official" FAST as it exists and as I use it in class/tests/etc. So this would not replace the FAST but rather become a variant to be used for things other than classification, pins, coins, etc.
If the delta had been a third of a second, I would have been willing to accept that. But Bill's actual data matches up to Gabe's mathematical estimation in a way that strongly suggests the delta will be significantly greater.
If frogs had wings I'd go back in time and do the surprise reload version from the beginning. But you guys have definitely convinced me that trying to change it now would be a mistake.