Im willing to second what Finsawyer said.
To determine the potential "pull", current should be what your mainly concerned with. In some of the most modern trends in auto development 42vdc systems seem to be the direction. It has been observed that low impeadence low voltage coils overheat, with continious cycling schedules. When a coil overheats it gains resistance lowering circuit current, weakening the solenoid pull. However assuming the resistance of an overheated low voltage coil, you could compensate with more input voltage, as ohm's law proves. I guess it all boils down to what your projected coil operating temp is.
I think what Electric Ed said about air gaps is pretty much right on square.
Just remember, if you have the perseverence to get thru reluctor and coil designs, you still have to contend with solidstate driving of the switch gear. Which really is not that tuff.
Have fun
-JW