Possible solution found.
I noticed that a tiny notch at the top margin of the bulb was not fully seated in the bracket (as seen in the attached pic). Not at all obvious even on close visual inspection - only discovered when I attempted to switch out the bulb as a last resort. So I secured the bulb, and did 2 test strips: #1 with Y/M both set to 100, then #2 with Y/M maxed out. No gels, and C=0 for both test strips. Both showed a heavy cyan shift. I'm going to try making some more test strips tonight with more reasonable Y/M settings around 70/60, etc and see what that shows. If this solves the issue, then everything may have stemmed from only a slightly malpositioned bulb, possibly loose enough to allow some movement between prints, resulting in unpredictable results. But we'll see tonight.