(Didn't finish writing this until after Howard's last post, but I'll leave this here for posterity.)
I'm going to take the opposite side of the radiator cap diagnosis.
First, while a radiator cap failure is possible, these bikes don't have a history of it. It's been the same part since 2002, and if Howard's 2008 had a failure, we'd probably have seen it multiple times on older models.
Second, I don't think the system would boil if only the radiator cap had failed. Some back-of-the-napkin math says that if the boiling point of 50-50 mix is 265?F at 15 psi (which is what the ST's cap is supposed to maintain) and the boiling point rises 3?F for each additional pound above atmospheric, the boiling point of the coolant with a completely-failed or outright missing radiator cap is 220?F. It would be reasonable to think that the ECM is programmed to start the fans at or just below that point (which, as others have observed, it does) to prevent boil-overs and engine damage in cases like that.
Third, the heat from the combusion process has to go somewhere. If the air around the radiator is stagnant (i.e., the fans don't run when they should), I could easily see the coolant accumulating the extra 45? required to get from the fans-on point to boiling, even under full pressure.
What would be really helpful is for Howard to start his bike and stay with it to observe what the temperature gauge does. If it reaches four bars and the fans don't run, there's the problem. I've done this experiment on my bike: it will idle with fans cycling and three bars pretty much indefinitely. Disconnect them and the temperature reaches four.
--Mark