SS3 time-varying biology on benchmark calculations testing

[e-perl-NOAA]

Hi all! If you attended Rick's preliminary NOAA internal presentation today, I said that I would start a GitHub discussion for those testing out this feature to post their results/questions/comments/etc.

Please refer to the presentation and notes as well as issue #191 and PR #640. There is a test model available in PR #640.

[Rick-Methot-NOAA]

Thank you for attending today.
I ran an additional test focused on using F%SPR40, rather than FMSY.
The old code and the new code both find the F that produces %SPR40 to be 0.107. This is expected because the search for F%SPR does not involve the S-R fxn.
When both code versions do the projection using that F, they get the same long-term SSB and catch because the S-R fxn used in the projection is that from the time series which is linked to the SSB/R at start year, not the benchmark SSB/R
The differences only occur in the benchmark calculations that depend on the SR curve. In this case the catch_MSY differs by 6.8% between the models. This difference shows up in the output quantities labelled SSB_SPR and Dead_Catch_SPR because these do the equilibrium calculations using the benchmark SR curve.