This morning, I reversed the direction of coolant flow through
the tubing, and ran some more tests. Mirabile dictu, now the
camera which reached a lower temperature was reversed: that is,
if the V-band camera got cooler yesterday, it was the I-band
camera which got cooler today.
So it really does look like a flow-rate problem.
I'll get a new pump and run yet another set of tests.
Michael