The direct coupled solid state (usually) power amplifiers do, with age, have a distressing tendency to fail in such a way as to dump (relatively) high voltage DC (up to and often at the rail voltage) to the XO and/or to the voice coil(s) of the hapless loudspeaker load(s) connected to them. So-called protection circuits... well... sometimes they work. Sometimes they don't.
I managed to fry two dump find,
cannon fodder loudspeakers with an amplifier (of similar provenance -- of course) in the above-mentioned state. The test speakers were single-driver
sans any XO, so the poor drivers had no chance. They
are electromagnets, but they're not designed to operate full time in the way of those things they use in scrap metal yards. Life for a voice coil asked to deal with continuous DC of substantial voltage (and current -- the DC resistance of the voice coil is typically pretty low) is typically nasty, brutish, and short.

The good news is that after the aforementioned debacle, I got into the habit of checking for DC at the outputs of foundling amplifiers prior to risking even
cannon fodder loudspeakers on them.
Anything that doesn't kill one makes one stronger.
I did actually manage to destroy one poor loudspeaker by carelessness with a properly functioning amplifier. I hooked my then-new Yamaha CA-610II to our old EV Wolverine fullrangers and played some symphonic thing rather loudly. The EV LS12 twincone is
not made for high power applications; it's a very sensitive driver and does
not require much power to produce high SPL. EV rated them for "25 watts", but that's sanguine. At any rate - things were sounding very good -- until one channel simply fell silent. Oops.
I still have that poor LS12 with its open voice coil in its original box in the basement. A
memento mori of sorts, you know?

A pair of LS12 drivers,
sans baffles, connected to an "Amp Camp Amplifier".