PMT or log amp linearity? (calibration beads)

Hello friends!
Continuing the calibration beads theme, spherotech sent the excel template 
(thanks!). BUT they use the title "PMT linearity QC record"... hmmm... in 
order for the machine's response to be linear, shouldn't both the PMT and 
log-amp(s) be linear? From Prof. shapiro's book, i understand the problem is 
usually the log amp(s) and not the PMT(s). Using beads, is it possible to 
determine if it's the one or the other? Not only that, but their excel 
template is in 256 channels, and their doc sheet states that if i used 1024 
channels, i have to convert to 256 in order to test it. Isn't that a 
needless "waste" of information, specially when i want to (semi)quantitate 
later on?

Uriel.