calibcant/theory.tex: Rework general derivations to remove G_1 (G_1p -> G_1)

calibcant/theory.tex: Rework general derivations to remove G_1 (G_1p -> G_1)

Values for the G constants are more easily calculated from the
equipartition theorem than through chaining G definitions.  I removed
G_1 entirely, and renamed G_1p -> G_1.

I also check my results against publishes sources.  It would have been
helpful if I had discovered burnham03 earlier :p.  They have a general
PSD formula, white and 1/f noise terms, and interference-based piezo
calibration.  Institutionalizing knowledge like this (vs. relying on
vast literature searches) is a good reason to consolidate work around
a small numer of open source tools.  Sigh.