COMPARISON OF MOTORS ON THIS WEBSITE
I'll try over time to compare the different motors for sale on this site.
I'll try and do it two ways:
1. by a technical comparison
2. by a riding experience comparison
1. techincal comparison:

The info above is from the torque graphs for each motor.
Just some notes on the table:
1. voltage refers to the voltage used during the testing to produce the torque
graphs (some of the motors can run on higher voltages
2. max. efficiency is peak efficiency from the torque graphs, the efficiencies
of motors changes depending on their rpm, shall try and go into more detail
later on.
3. max. torque is the max. torque possible just before the motor stalls, each
motors torque has been converted for it to be in a 26" rim. Results for P2a
motor a bit difficult as the torque graph didn't say which rim size it was
calibrated for.
4. max rpm = this is the max speed of rotation of the motor when its under no
load (i.e. the same as having the motor run flat out when the wheel is in the
air, so top speed of bike cannot be worked out from this figure it will give an
over-estimate)
4. max. Pout = maximum power out the motor can deliver to the road
5. max Pin = the maximum power the motor/controller uses ( Pout and Pin are
directly related according to efficiency of motor at any given rpm: I forget the
formula!! will look it up)
6. top speed in a 26" rim is taken from road test rides not from max. rpm
of motor (in no load situation).
Although these figures are very accurate at the time the test was done, in real
life riding conditions they can only be taken as a guide. For example on a 408
motor using a 20amp max. controller at 48volts I will get a maximum of 960watts
power input according to a drain brain meter. In the test by a torque machine it
was tested to be 860watts, the difference may be due to the maximum current of
each controller may be slightly different and not exactly 20amps. So these
figures can only be taken as an " as accurate as possible" technical
summary.
Some Notes: on each motor
408: can be run on higher voltages (up to 72volts) but motor tends to get quite warm/hot on 72volts when under load, not recommended unless your prepared to keep an eye on motor temperature. Haven't tried regen. type controller with this motor as yet but should be ok.
gl-1: specs are basically identical to 408 motor. Haven't tried running on higher voltages as yet but would probably find temperature of motor will get hot/warm similar to the 408 but I'm yet to try it. Regen braking function on controller with this motor works very well. In a 20" rim it adequate to stop bicycle without using brakes, in fact the regen is quite strong in a 20" rim, yet to try on a 26" rim.
P2-b: should be a very good motor but haven't finished testing as yet.
408/4012 dual speed crystalyte: differences in torque compared to gl-dual
from each winding is greater in this motor than gl-dual.
gl-dual: seems the difference between windings is not huge only 15% but
yet to test to see how it performs on a bike.Have started testing on the
autoswitching controller, it changes windings used based on speed of motor,
still testing controller with it.
P2a: motor tested was 24volt version, it can also be run on 36volts but I haven't had a torque graph made of 24v motor running on 36v as yet
x5: above results were for an x5 on 36volts, it is designed to be able
to run fine on 72volts (some people are using it on 84volts using standard
36v-72v controller which seems to handle 84volts no probs. Controller cant
handle over 100volts). Hoping to get some torque graphs of it running on
different voltages in near future.
Because this motor is really designed for higher voltages the results here only
give a very narrow view of what its capable of.
P3: a note on the torque graph for P3. The initial graph I was given was from a supplier who initially sold me these motors. I had another test done by an independent company but the motor had some problem, the no load current was 4.5amps indicating there was some problem inside the motor (the no load amps should be less than 1 amp), so unfortunately I cant use that graph. Will have to get another motor tested at some time, so till then I can only go by the graph I was originally given by supplier. It states 75% max. efficiency I would be inclined to think the max. efficiency should be less maybe 70% given the metal gear train in the motor but will have to wait till I can get another P3 tested to confirm if 75% max efficiency is correct/accurate. Results show very high torque for this motor as expected. The motor can be used on 36v or 48v.Results above are for 36v only hopefully will get a chance to get some torque graph made on 48volts also.
beware of spray ebike company: more info