Most amplifiers would run for many seconds if the those designers eschewed AC detection power off circuitry. Fortunately, such circuitry has been present in most quality HiFi gear since the 1970s.
Long run time on power off is not a sign of anything good or superior. It merely highlights the lack of adequately complete protection circuitry. Some amplifiers with non-symmetrical current draw from the PSU rails will exhibit considerable offset fluctuations as the unpowered rails collapse differentially in such a situation. Competent design includes either front end muting, or speaker disconnection upon AC cycle loss. This is normally monitored direct, from a half wave rectification of a secondary tapping.
In the case of the MC2500, they likely figured relay disconnection was unnecessary due to the output transformers and instead of muting the input stage on AC dropout, they saved a few cents and simply didn't bother with any power-off muting at all. According to the schematic, they spent a whole lot of time and effort on their turn-on delay and power-guard circuitry and simply ignored turn-off muting.
It also appears that they omitted the virtually bog-standard bleeder resistors over the main filter capacitors, installed to safely and predictably discharge the otherwise high voltages present after turn off.
Very slack engineering on a TOTL 500w/ch monster amplifier IMO.