Not in 1994 they didn't.
SDRs are called for when there is a need to operate on a wide range of frequencies dynamically. Emergency beacons in 1994 required exactly the opposite: they had to operate on a fixed frequency to within a very high tolerance for frequency stability, and very long term reliability.
Your source is from 2016. Nowadays we use several channels in the designated emergency band 406.000 MHz to 406.100 MHz, and modern beacons may have to be compatible with legacy receivers and spacecraft. There are currently four fixed frequencies in use, requiring switching between them in normal operation. That makes SDRs the preferred technology in modern beacons. The frequency drift tolerances are still the same for each discrete channel, and none of that behavior is available for external control. There is no way to make the radio not heard on the two channels in use in 1994.
Sorry, as usual you don't get to map modern practice back onto history and pretend you understand anything about it. You cited two sources: one a Wikipedia article giving the history of software-defined radios, and the other a manfucturer's spec sheet for modern SDRs certified for use in beacons. Had you actually attempted to assimilate and combine the information from those two sources instead of merely trying to beat your critics over the head with them, you would have seen why your claim fails.
You claimed they had been sabotaged by having been tuned away from the proper frequency. That is not possible.
No.
Normal testing is simply to ensure the battery charge. In the 1990s, the lengthier (e.g., annual) test to which you refer requires a test rig, such that the EPIRBs would have to be removed and taken ashore.
No.
As you agree, Lehtola was not an expert in EPIRBs and did not describe them accurately to the reporter who asked.
Because the EPIRBs had not been switched on. This was determined to everyone's satisfaction but yours. And since you don't understand how they work, we know the cause of that confusion.