Conclusions: 1) The added offsets didn't seem to make a large difference. 2) These don't looks like (typical) DAC glitches - The character of the prominent glitches around 50 Hz at LHO at the moment are medium-loud glitches separated by seconds to tens of seconds that appear as multiple fringe-like glitches that peak above the detector noise for about 2 seconds.

Spurred by Jeff's addition of offsets to ETMX and ETMY to possibly avoid DAC glitching in 44866, I had a look at the character of the LHO glitches before and after 10-28 20:00 UTC when the offsets were temporarily added. I did not see any significant change in the character or rate of the glitches. Figure 1 shows the glitch rates around 20 UTC, they don't change much (the little colored dots would have dropped around 20 UTC and they didn't). Figure 2 shows a spectrogram before the offsets were added. Figure 3 shows a spectrogram after the offsets were added (I looked at a few times before and after). 

However, when looking at the glitches in these spectrograms I noticed that there are brief medium-loud glitches separated by seconds to tens of seconds and that the glitches appear to have little fringes at their tops. Looking more closely with omega scans (figures 4,5) there are multiple fringes that peak above the noise for about 2 seconds each time. But not all the glitches look like this and not all of them are as clear as these examples (I don't know why, maybe because when the cross strong lines our whitening filters mess them up or maybe they are really messy).  

No idea what's going on yet. Maybe some suspended optic is given a brief kick and is then damped quickly (seems unlikely?) or maybe something is moving a lot and its coupling into darm is modulated such that it only appears briefly and separated by tens of seconds. Suspicions welcome. Detchar folks could try lining these up with suspension DAMP L channels or looking for excursions in, say, alignment channels at the same time.