Allow P2 > 1.8 for OR's and P2 > 2.8 for ER's in final review

[taldcroft]

See the like-named section in https://occweb.cfa.harvard.edu/twiki/bin/view/Aspect/StarWorkingGroupMeeting2021x01x27.

This new policy should be reflected in the starcheck output somehow.

[jeanconn]

Since we left the "if it was > 2 or > 3 in prelim" I think this just ends up being a comment if you think we need it.

[taldcroft]

After 6 months I will have forgotten about this policy, so yes it does need to be written.

[jeanconn]

Might be useful to come up with a similar threshold for the guide star metric and at what level we think creep-away is a good idea.