The method I listed above was to square the rods not the frame, however just the way the SD is constructed if the frame gets bent the rods will follow... from what I saw in your first picture there is a fairly large deviation for a 3 inch part if it was the frame that was bent it your printer would be wobbling at least a half inch on three legs, in the second picture you are checking the rods from the bottom sheet metal of the printer... because of the design of the printer the floor may bow down so checking the rods against this will not be accurate or useful. I think you will find that the settings will have the most impact on how the holes come out... when ever I print holes I always undersize them and ream them out afterwards.

That would help to square them up, as my drill press is indeed perpendicular, but it would be better if my parts came out as per design without having to "fix" them.  I am printing up this test block to see if I can determine where the error is.  The design is a 50mm square and high unit with 5mm thick walls.

Are you sure it is off in both x and y... you can adjust the x axis by moving the white bearing blocks up or down as needed. I would check the machine for square before relying on printed holes with threaded rod as an indicator....also easily accomplished by placing a machinists square or other on the bed with the vertical edge resting on the extruder motor shaft then move bed up and down but do this after the bed is leveled very good across x for checking x and leveled very good across y for checking y.

I am only home one day of the week, so I have not been able to fully investigate the phenomenon enough to be certain of where the issue  lies.  However, since holes in parts and squareness of pieces are what I am trying to produce, the results are for me far more important than whether the frame is square or not, albiet I understand that indeed this is probably the root of the problem.

All of the axis combinations could be adjusted for perpendicularity, some easier than others... So which axis's are out of square? x to y x to z or y to z

It seems that the Z is not perpindicular to the x and y.  I can't see any way to adjust since the guide rods are held in relative position by holes drilled into the frame  By the way, this is not a warpage issue since the part stayed completely flat and stuck to the glass for the entire build.