Changes to default --status_{dt,nt} behavior

[Rhys Ulerich]

Hello,

First off, I'm going to start using the mailing list a bit more for discussion.

Second off, the actual point of discussion related to https://red.ices.utexas.edu/projects/suzerain/repository/revisions/44243...

For apps/perfect/driver.cpp the default restart_dt duration was 8 times per flow through, the statistics 4 times per restart, and the status 4 times per statistics.  That meant a status message 128 times per flow through.  I increased this to status 16 times per statistics for a status message 512 time per flow through.  Here "flow through" is loosely defined to be an approximate bulk velocity for channels or the freestream velocity for boundary layers.
  
To coincide with this increase, I removed the default --status_nt=100 line appearing within the jobscripts.  Folks wanting status more often can explicitly provide STATUS=--status_nt=100 at job submission time.  That means, unless you override it, the status messages are now equispaced and driven by simulation time rather than by CFL-related considerations.  I made the change because --status_nt=100 trampled all over --status_dt="flowthrough/512" preventing it from working well.
   
Lastly, I added a TODO to apps/reacting/driver.cpp suggesting that it be updated to provide default restart, statistics, and status behavior when run in a boundary layer configuration.

Give a shout if you're unsure how this might impact you.

- Rhys