Thanks, everyone, for your advice.

If I spot the 100 kHz crystal oscillator at 3.7 MHz, I have no more
than 2.5 kHz of error across the band, which is a bit high, but
acceptable, so I’ll leave it alone for now. If I do attempt to reduce
the dial error, I’ll try bending the tabs on the variable cap. I’ll
first attempt to do this while the LMO is still installed in the
receiver. This will require a dental mirror and pick, but may just
work.

-Dave Drumheller, K3WQ