I don't think there is any real technical reason that 'shorter than 5 minute development times' will not work, but it seems more a practical one of how consistent you can be at timing and stopping development. A 20 second variation at 6 minutes is a 5.5% variation. The same timing error for a 3 minute time would be 11%. If you are tray developing and using and acid stop, I imagine your potential timing error would be small.
As an aside, there was a thread on the Bostick and Sullivan alt-photo list a few months ago about using amidol as a developer for platinum negatives (Terry King says this is what Frederick Evans used). His times for FP-4 were about 2 minutes if I remember correctly! So if you're in a hurry, there you go!