+(*)Add OBC_TS_RESERVOIR_INIT_BUG runtime param

[Hallberg-NOAA]

The code that was previously wrapped by the runtime parameter OBC_RESERVOIR_INIT_BUG actually included two separate bugs in the order with which the tracer reservoirs were initialized. This commit adds the new runtime parameter OBC_TS_RESERVOIR_INIT_BUG that wraps only the point where temperature and salinity in the reservoirs are set with new runs, and not the bug that sets when the other tracers are initialized, which is still selected with OBC_RESERVOIR_INIT_BUG. Although there are cases where the answers could be changed by the value of OBC_TS_RESERVOIR_INIT_BUG, this only applies when REGRID_ACCELERATE_INIT is true, which appears to be uncommon. It is even possible that there are no cases with active OBCs and tracer reservoirs that also use REGRID_ACCELERATE_INIT=True, in which case OBC_TS_RESERVOIR_INIT_BUG could be rapidly obsoleted, whereas there are cases (like the loop_current test case) where the value of OBC_RESERVOIR_INIT_BUG does change answers. All answers in test cases are bitwise identical, but there could be some cases where this new parameter changes answers.

[theresa-cordero]

I have not tested these changes, they are clear enough that I don't think the usual testing is needed, but I support separating the one bug flag into two this way.

I also support obsoleting OBC_TS_RESERVOIR_INIT_BUG if possible while maintaining the other bug flag (if needed).

[theresa-cordero]

If @herrwang0 is happy with this change and how it will fit with #1139 then I will approve this PR.

[Hallberg-NOAA]

This PR has passed pipeline testing at https://gitlab.gfdl.noaa.gov/ogrp/mom6ci/MOM6/-/pipelines/31256 with the expected warnings about new runtime parameters.