It sounds like this audio glitch is some sort of full-scale-ish dc squarewave. Depending on the native filtering present on the amplifier, the result will be some degree of "pop" from the initial pulse of the signal. If there is not enough dc-filtering, the "pop" will be accompanied by some amount of infrasonic pulse (perhaps, the character of a "gunshot"). If any of the analog stages in the chain are getting overloaded from such a signal, they may contribute their own additional break-up modes to the audio result, as well.
As for why it occurs, maybe there is some errata in certain decoding chips (maybe not even isolated to a particular chip family, rather an implementation version that was in effect on some manufacturing lot, one week). The errata causes a disruption in decoding when the right conditions are present to trigger it. Normally, this would not be a problem (as a digital decoding spec should certainly have some built-in tolerance to corrupt data or intermittently poor signal conditions). So, perhaps this is suggesting a bad failsafe mode in the DTS Master algorithm, itself? When corrupt data or temporary interruption of stream is encountered, the worst that should happen is a quiet click or a repeated sound sample (if you can hear it, at all). If the decoder issues a fullscale "pop" or has to reset from a crash which causes a nasty "pop" transient, that is very bad practice. Somewhere along the line, a failsafe mode is not behaving as it should.
A drop-out in hdmi encryption should not be counted out, as well.
I don't think this is the fault of any one
party (which makes it that much more difficult to get action, of course). I DO think there are specific parties that should be involved to investigate this and make any corrections they can in their respective arenas.
DTS should be investigating their failsafe modes and verify that their licensees are producing properly compliant chips (such that these failsafe modes are working properly in these chips). The data stream could be good/bad/intermittent, but under no circumstances is it appropriate for a fullscale "pop" to be the resulting sound.
Similarly, manufacturers should be verifying that the DTS chips they are using are working according to spec (especially, regarding the failsafe modes from bad signals or crashes).
Disc mastering houses should be verifying that they are indeed encoding DTS legal streams, in the moments where these "pops" are recurring.
Jesse Jackson may require some charity grant, as well. I have not heard him weigh in on this, just yet.