A normal regulator can run MUCH hotter, and will be worse as the voltage increases. Anything over 2s will be bad, which is why the ESC BECs should be disabled over 2s. Also, most of the types at radioShack are only rated for 1-1.5A anyway.

Hmm. What about a regulator that had over-kill ratings, I'd assume it wouldnt get as hot since the input voltage isnt getting anywhere near its rated spec? Just thinking out loud really- if the MMM bec goes tits up, the whole esc dies, so thats not an issue, in that respect.

The formula for power loss in a linear BEC is: (V_batt - V_bec) * A_bec.
- For a 4s pack (14.8v) and a servo load of 2A, the loss is: (14.8-6)*2=17.6 watts.
- For a 6s pack (22.2v) and a servo load of 2A, the loss is: (22.2-6)*2=32.4 watts.

Oh crikey, thats not quite what I had hoped for.

Along the lines of your previous suggestion then, what about using some kind of component that was only rated for 8v say, that would blow if the BEC failed and tried to send full pack voltage to the Rx and servos? Same idea as a normal fuse, but one that was voltage sensitive rather than current senitive. Or am I just being dense again?...

Yeah, a voltage sensing switch is the better idea, but the circuit is a little more complex than a simple regulator. I had brought up the idea a while back, but never went anywhere with it. Ideally, this type of protection should be integrated as part of the BEC circuit. It would make the package bigger, but worth it IMO.