Clean up of tables code

Clean up of tables code

This commit is not intended to change any correct and useful
functionality.

Removed various debug code. This is what unit tests are for. If we
want the ability for a user to dump the tables we've generated, we can
do that, but not as part of the run setup code. We pass fewer rarely
used parameters now.

Removed bGenTab. If we've read tables in, then we already have the
metadata for any tables that we will later need to generate. Only if
we have read no tables do we need to set the metadata to the
default. David introduced bReadTab and bGenTab in 4fd28a39ea32 to do
this, but it was more complicated than it needed to be. Berk changed
the logic in 8e6bc73800ab, and it doesn't seem right with respect to
the overall behavior (or specifics of Ewald user table types). Renamed
bReadTab to useUserTable to clarify. ETAB_USER is now used a few more
times, where probably it should have been used all along.

Made initialization of the x field of t_tabledata something done in
the various code that fillins table rather than init_table, since it
clarifies other things and the x field was in several cases being
over-written anyway.

Moved Buckingham handling code into the generation branch, which was
the only one where it did anything. Eliminated unused scale_exp field
of t_forcedata.

Moved assignments to nx0 variables now that the role of bReadTab is
more clear.

Eliminated useless nx variables.

Change-Id: Icdd88612b28ce278bebac6c138445d90145ef513