A cordless toaster eh? What's your toast time (runtime)?
JRH: I attempted to test output variance with fluctuating input voltage. I hooked the output of my 14.8v PS to an external linear regulator (with pass transistor). I used an LM7812 and set it up in adjustment mode. Then, I put the adjustment circuit on a rotary switch so it alternated between 12v and ~14v as fast as I rotated the knob. This was the only way I could think of simulating a quickly changing input voltage. I used my 2.5 ohm resistor as a load to pull ~2A from the BEC.
The output did change, but by just 20-30mV. However, I'm not sure my linear regulator setup was able to change voltage as fast as needed. Even still, I'm at the mercy of my voltmeter sampling rate. Soooo, I hooked the BEC output to the same load and a Spektrum receiver/servo and tried it again to see if the receiver would glitch. Nothing; it worked just fine.
Do you know if the people seeing the problem were using ESC braking? Depending on the setup, the back-EMF from hard braking could be generating voltage spikes exceeding the BECs input rating. Something to think about. Most other BECs on the market are rated up to 35-40v, so there is more headroom if this is the case.