YellowJacket said:
Mesa 'claims' that 8 ohm output with a 16 ohm cab is a correct mismatch. An 'incorrect' mismatch would be an 8 ohm output with a 4 ohm cab.
Yes, they claim that. That doesn't make it true.
Generally speaking, one shouldn't mismatch. If you have to, going numerically lower is safer. So your second example is generally safe (it's how all those BF and SF Fender extension speaker jacks worked). Going the other way, not so much. Lots and lots of discussion on this out there, try RG Keen's tube amp FAQ.
The two cooked amps both experienced arcing between the plate and the heater contacts on one or more power tube sockets (pins two and three). That dumped the plate voltage into the heater circuit and burned the board. Oddly, Mesa used flameproof resistors everywhere EXCEPT the heater supplies, pretty much ensuring incendiary results.
The cause of the arcing was flyback, and that was almost certainly due to upward impedance mismatch.
I imagine they were just using up old inventory on the OTs, or maybe didn't see the need to design a new one. But when these amps came out, I'd think most of the target market who owned a 412 owned a Marshall cab, so they had to at least pay lip service to 16-ohm operation. And it'd probably have worked fine, if these were say jazz or surf amps.
Anyway, Mesa has a history of releasing products before they're really ready...just about everything they've ever made underwent numerous revisions, usually not readily apparent externally. The power supply on the early DRs is really poorly designed in general - for example, the switching system takes power from the bias circuit, and just knocks the voltage down with a giant resistor.