Current implementation of CHE may have unintended or undocumented consequences

[ilyamandel]

Some of the code for treating CHE stars may have unintended consequences, or intended but perhaps not fully documented consequences. Two examples:

• the stellar rotation angular velocity is set to orbital angular velocity for any binary after CE or MT if CHE is enabled; this could have unintended consequences when treating tides;
• CHE binaries always circularise with a conservation of angular momentum, ignoring the usual flag settings for whether circularisation should happen at mass transfer, and, if so, whether angular momentum should be conserved.

[As a sidenote, the function BaseBinaryStar::CalculateMassTransfer(), via BaseBinaryStar::InitialiseMassTransfer(), somewhat inconsistently amends the orbital parameters -- and even stellar masses for CHE stars -- on circularisation, whereas usually the future orbital updates on mass transfer are only recorded and delayed until ResolveMassChanges is called; this can manifest in unexpected inconsistencies in the code.]

[ilyamandel]

Implementing tides properly -- #210 -- would address this.