ci_hsc has a test that verifies that extendedness as measured on coadds broadly agrees with the star selection done for PSF estimation on individual frames. This tests a bunch of stuff, including aperture corrections on the coadds and propagation of flags from visits to coadds.

It doesn't test that aperture correction vs. extendedness logic is correct in processCcd.py, but just copying this test to the appropriate validation function in ci_hsc should do the trick. This is currently broken, but should be fixed in DM-5877.