Updates for failing tests

[mnlevy1981]

1. with DO_SKEB=True by default, we need to turn it off in the no_stoch_physics test
2. In order for ERI tests to pass, need stochini=True for hybrid runs

[mnlevy1981]

The good news is that the no_stoch_physics test passes now. The ERI test is still failing; ref2 is looking for ./INPUT/ocn_stoch.res.nc rather than the reference file. I wonder if the problem is with a branch run rather than hybrid? Switching the logic and testing that, but this might require a MOM6 update to make sure we set stoch_restfile correctly for hybrid and / or branch runs...

[mnlevy1981]

I have verified that ERI.TL319_t232.G_JRA.derecho_intel.cice-default and SMS.TL319_t232.G_JRA.derecho_intel.mom-no_stoch_physics tests pass; @fischer-ncar reported those (and an ERI test with a B compset that has the same issue as the G compset) as well as an MCC test that is not addressed in this PR... that will need to get fixed in a later tag.

I have not run aux_mom or pr_mom, just the one-off tests.

[alperaltuntas]

Tests are ongoing. I'll merge this PR for now so that alpha tests can start too.