Entertainment Center: TVs, Stereos, VCRs and DVDs - Mitsubishi CS-2715R horizontal fixed! (almost)

A while ago, I complained about my 27" Mits CS-2715R
having a shrunken horizontal scan--4" of black bars on both
sides of the screen with full vertical deflection.

Originally it started as 2 vertical bands, later progressing
to pincushioned bands with a hump in the middle of each
side of the screen; then, ultimately the TV died completely.

Found the problem in the Pincushion stuff--there was a cold
solder joint on 2 manually reworked back-to-back zener
diodes that serve as a 12V reference to the PCC IC. The
12V comes off the 16V from a separate transformer.
There's a series 0.33ohm power resistor
running off the 16V which opened up as well. I can only
surmise with the resistor opening up, there was enough
drop down to 12V, so I didn't catch this the first time.

The manual rework was covered up under a glob of adhesive
caulk (or whaterver they use to tack down wires, caps, etc
on the opposite end of the deflection board, so this was
a bit elusive.

I've replaced 36 leaking/bad electrolytics between the
defl and signal boards. What a mess of electrolyte!

So now, the TV is considered 'fixed', but vertical linearity
seems to be off. A circle test pattern looks somewhat
like an egg with the fatter section on the lower half of
the screen.

I've tried messing with the vertical height/linearity pots with
no avail. If I set it so the circle looks like a circle, the entire
picture is shifted up with a 2" blank band on the bottom of
the screen, while maintaining the center of the test pattern
at the middle of the screen

How much nonlinearity is considered acceptable or is lousy linearity something common on lowend TV's? Watching
regular TV program looks "OK". But with the test pattern,
it's pretty evident.

ideas?

btw, Thanks to Teratum, et al for all insights thus far!

aaron