you need to consider driver excursion as well!
IMO you need to balance AT LEAST max SPL, magnitude response and excursion (port air velocity would be good too to prevent "farting")
I haven't designed anything of my own, just simulated known designs with different drivers and some of my own ideas (only CB and reflex so far!)
my method (corrections and improvements warmly welcome! :) ):
1. adjust box volume (and tuning freq for reflex) if unhappy with Winisd's suggestion 2. check freq response 3. set signal to half rated RMS power 4.
examine excursion plot -> if it's at half xmax, you can mess with
box size or eq (linkwitz transform) to get the LF response the way you want
it; if excursion is close to or exceeding the limit there is something
wrong! at the very low end between 20 - 40 you will usually run out of excursion
anyway, so you can set a HPF to limit excursion there. Above 40 Hz, you can actually decrease excursion by tuning the cabinet higher (port resonance decreases driver excursion!!!). 5. recheck magnitude response and SPL; if the curve looks good, check if there is excursion to spare -> if yes increase signal power until you reach excursion limit; That gives you max input power and SPL max for your current design 6. if not happy with freq response, you can readjust box volume, tuning freq and EQ/filters
all these are interdependent. Bigger box also decreases cone control -> increasing excursion What happens depends on the individual driver.
Try simulating many drivers with a certain set box volume. You will see the difference between the different drivers, some will be impossible to get acceptable results with......   
IMO one way to recognize a potentially suitable driver for your design goal is if WINISD automatically comes up with a box volume close to your design goal - you shouldn't have to fiddle around too much to get a good result.
|