6v6gt:
The “on all the time” and ~0.5 voltage drop sounds like the heater is being powered via the mosfer body diode, which would indicate a wiring error (although I’m sure you would have checked that). Odd is also that the voltage at the middle of a potential divider appears to be independent of the voltage across it, whether 0 or 5 volts.
Maybe for testing, use a simple sketch which explicitly sets D10 to either high or low instead of sketch you have posted.
I'll do the testing, thank you 6v6gt!