We experimented with non-sidereal tracking at quite high delta's (1800 to 5000"/hour) from sidereal and found that pointing errors were quite large, up to around 500 pixels.  There appears to be a cycling effect here - the errors accumulate, reach a threshold around 500 pixels off in RA, then drop down to a few tens of pixels before accumulating again.  This takes place over 5-6 images of the same field.  Given that this is not a normal use case for the telescope this is not something that *requires* a fix, but it may be interesting to consider why the telescope behaves this way.

Along the top edge of the new wireless keyboard for observer3 there is a row of small and apparently innocuous special-purpose buttons.  However, one of them is marked with the power symbol and "PC".  If you inadvertently press this button by, for example, picking up the keyboard by its top edge, it will turn off all the monitors and render the computer unreachable from observer2.  No amount of panic and pressing the button again achieves anything.  The light on the PC cabinet stays lit.  There may be more direct means to bring everything back, but all we could figure was to hard restart observer3 which thankfully brought it back from zombieland.  This is clearly not the best button to have so easily in reach.  I'd pry it off with a screwdriver and destroy the switch underneath, but there are probably more graceful ways of disabling this short cut to an observer heart attack.