Differing IMS quadrupoles between acc and original TRANSOPTR model

IMS quadrupole lengths and aperture seem to differ between acc/ database (sequence ims_db6.xml) and beam physics TRANSOPTR model used for theory tune computation: http://lin12.triumf.ca/optr/ISAC/Sep-LEBT/matlab/sy.f

Specifically, see IMS:Q11 and Q12 (each representing two quads), which differ between versions.

I'm tagging @baartman and @spencerkiy for awareness. I'll handle fixing this, but am tracking bugs/finds as I go along!