If they're rebuildables, have them apart & clean all the gunk out, then inspect the commutator & brushes & feel/look for wear/play in the bearings. Twiddle the spindle to get an idea of the condition of the magnets (More "cogging" usually means more power).
Push & pull the spindle, there should be a small amount of clearance both ways but not excessive & not just in one direction.
Check the brushes move smoothly in their holders & don't have excessive clearance or bind.
Check the wiring & solder joints (I just replace that lot without even looking, there aren't many people who can solder properly).
Check the suppressor capacitors are present & in good condition (Some motors have these as SMD chips on a small PCB on the back of the motor).
Invest in a comm lathe - you can't do without if you're reclaiming motors! - out of round commutator is the biggest cause of poor performance & is the most likely thing to be wrong with a motor, pre-run or even new.
Don't buy pre-raced motors, they've usually already been machined until the comm is wafer-thin & can't be trued
For testing, I usually run them up on a charged battery & listen in the first instance - you can instantly tell if the bearings or comm are worn as they have a distinct clattery sound & don't sound as fast as they should - compare the sound with that of a known good motor. Look at the brush end, excessive sparking is a sure sign of a naff comm.
Put gently pressure on the end of each brush in turn & see if that makes the motor run faster (Weak springs or naff comm).
Shake the motor with it running, sideways & end-to-end to listen for clunking (Excessive bearing wear or poor endfloat shimming)
To prove the motor, I usually try it in a car that I already know how it runs with a good motor in it, & compare the expected performance with the unknown motor.
Lastly, after running in the car, make a note of how warm the battery/wiring is getting, & how quickly the pack went flat in comparison with normal. Obviously tuned motors are going to do that more quickly anyway, but excessive comm sparking, worn bearings, feeble magnets or poor winding insulation will warm things up despite an apparent lack of performance.
Electronic test gear probably won't tell you much, as it's the performance at varying loads, speeds & conditions you're interested in, & bench testing won't be able to mimic use in the car.
You can measure winding resistance (& inductance if your meter will do it), sometimes this kind of testing might show up shorted or broken windings, but usually if they're that bad, you can tell by the general look of them.
Resistance testing can also tell you if the brush springs are weak or the brushes sticking in their holders.
If your DVM can measure capacitance, you can check the suppressors, but to be honest, if you're going to remove them to test them, you might as well just replace them anyway.
You can measure unloaded running current to get an idea of how tuned the motor is, but this can get confused with how mechanically ill it is too.
I find looking, listening & trying the motor in a car is far more productive than using test gear on it.
At this point, it's worth noting that if you have several motors made by the same manufacturer, you can sometimes swap parts about to get a couple of runners out of a pile of dead ones (You can do it with most if you don't mind a bit of shimming/modding)
