lets start with some general information so that people can decide how to use a motolyzer as somewhat of a dyno.

General rules of thumb:
Decent stock motors end up having an efficiency around 60-70%. Lets say 70 to generalize.
efficency is power in versus power out.
torque and efficiency for each motor is a curve not a straight line.
For example torque is a bell curve and so is efficiency
power (rpm and torque) looks kind of like a nike swoosh
rpm is a function of torque. Without torque you get no rpm

you can determine the motors efficiency and relative torque with a motolyzer
there are 2 curves for torque. 1 for max loaded torque and 1 for peak efficiency. Max loaded would be slightly more current and rpm. You cannot simulate max with an unloaded motor but you can slightly advance it to approximate it. You can see it in lap times that do not overheat the motor.

the math
do an xy plot.
30 degrees = 10000 rpm and 4 amps
40 degrees =1500 rpm and 5 amps
50 degrees =18000 rpm and 8 amps

there is a torque constant for each motor you are recording but not directly. You are substituting rpm for torque beacause you cannot measure torque but current is available to help.

rpm/amp is the torque constant and its what you just plotted.

you will find a bell curve and the peak of it will be peak efficiency. If you slightly advance it it will be peak power. Peak power generates peak heat. So slightly advance the timing of your peak efficiency. Or run it at peak efficiency to keep it cool and adjust the fdr for heat then slightly advance and back off a hair on fdr.

in 17.5-21.5-25.5 the torque constant will have a steady slope up to peak efficiency then it will be flat for a couple thousand rpm and then the slope will go negative. The peak is around a zero slope.

If you found a timing (for an fdr) that is the max heat you want that motor ever to run you know the amps/rpm (torque constant) or thermal time constant. Once you have enough slopes plotted for a series of fdrs you know the ability of your motor to dissipate heat. This can be a line through the rpm/torque slopes you created that limits tells you how far you can advance timing above peak efficiency for an fdr. It will work for 3 or 4 fdrs (depends on slope up to zero and down from zero) if those fdrs are within the bell curve for peak efficiency. If the thermal time constant or peak power curve is too complicated to understand before you try this then just get peak efficiency and FDR the car to the peak efficiency you calculated then go to slightly lower fdrs in increments until the car gets hotter.

the motor will get slightly hotter if you dont reach the rpm you calculated but it will be slow and potentially much hotter if you exceed the rpm you calculated. Amps/rpm usually dives very quickly at higher rpm on most motors but there are a few motors that the slope down (negative) is about the same as the slope up (positive) but that is pretty rare.

Most motors have a shallow slope to 70% and a sharp slope down. The motor I tested that is the exception that is shallow up and shallow down was the r1 21.5 v21. This sounds great and it is but it means the thermals are more complicated. If you have a peaky motor that drops quickly after peak its easier to calculate the max rpm you will ever need and keep it below the negative slope and not be as concerned about the thermal dissipation. Why-because the motor turns into a dog abruptly which means it also generates a ton of heat quickly which is easier to observe than a slow taper up and down.

the motolyzer will not tell you torque or output power in units of measurement. If you are in the ballpark for efficiency current in has a linear relationship to torque out. Using the xy plot will give you extremely accurate numbers for what it is. You can compare one setting to another. You can compare one motor to another (unloaded). If you find peak efficiency or peak torque using the xy and someone told you what that peak torque equals on a motor dyno you could apply that same torque to other peaks on an xy plot.

in light of recent threads i would say that it is also a bad idea to compare values from 1 motolyzer to another motolyzer. I would think that comparing curves generated from one motolyzer to another could be compared if your comparisons are relative measurements based on the peak of each curve generated by different analyzers. I would assume that the differences from one analyzer brand to another is at least a problem in the current resolution and calibration. Absolute measurements like 5.8 is the peak efficiency on one maybe be 6 on another. But if you find peakE and then compare changes based on plus .2 amps to get peak power from the center point of efficiency the differences in calibration from 1 to another should be minimal.

Let me try describing it one more time.

1-set your end bell to 45.
2-use graph paper where each vertical is .5 amps and each horizontal is 1000 rpm
3-Run the motor in the test that tells you rpm and current
4-plot 10 points at 10 rpms (current and rpm)
5-start a new curve and repeat all steps at 50 degrees or 46 degrees (your choice)
6-repeat step 5 with different timing marks
7-interpret the data (40 degrees)
1-for example the slope up to peak below 40 degrees (on your graph paper) will be an additional 1 amp per 1000 rpm. (For example)
2-peak efficiency (0 slope) will be .5 amps per 1000 (but this peak is at 9500 rpm)
3-slope down or negative slope will be 1.5 amps per 1000
8-interpret the data (45 degrees)
1-same as 7.1
2-peak efficiency is at 10000 rpm
3-same as 7.3
9-interpret the data (46 degrees)
1-it takes more current per 1000 rpm than 7.1
2-peak efficiency is at 11000 rpm
3-same as 7.3

What did you learn?
1-46 degrees takes away low rpm efficiency (comparing 1 curve to the next)
2-45 degrees retains all low end torque but you peak at 10000 rpm (this is a great curve for short tracks with lots of accel and deceleration)
3-the slope down from peak is almost always the same but adding timing moves the efficiency peak to higher rpm
4-you dont ever want to cross over the negative slope side of any of your curves so the max rpm you need on a track should never exceed sit outside this rpm
5-some of the negative slopes are sharp and some are shallow.
6-calculate the average rpm you need based on fdr and roll out. If you negative slope was shallow set the timing and fdr so that you rpm needed is at the average rpm you want. If the slope is sharp set the fdr to a little less but leaving timing alone so that your max rpm stays inside the curve you drew.
7-all of these curves can be compared to other motors and they will be accurate even if you dont know the exact power or torque

you dont have to plot 3 points. It could be 10 points and they can be in 1 degree increments or whatever you like but the slope down from the peak will be consistent. The slope down (negative) will look like more current in and less rpm out.

For example the points you plotted before the 0 slope. 45 degrees=2 amps per 1000 rpm, 1 amp per 1000, 3 amps per 1000.

Ill try to do a video but maybe my clarification helps.

To create the x-y plot here is an example. Im not a video guy. The goal is to find the peak of the motor. Substitute amps for torque. There are 3 points here on the x-y plot. Each has an amp draw and an rpm. It will look identical to the underlying torque curve except you are substituting amps for N/mm.

Once you can make the xy plot. This process gets you the unloaded peak power which is the best a motolyzer can do.

do it over an over again to find the highest peak (point 2 in mine) by adjusting timing.
what you just found is peak power with neutral plane or sensor timing efficiency. In most cases you can leave the timing alone after this.

Since the esc in a motolyzer is power limited you cannot use its ability to measure rpm and current on the chassis because the load and friction introduced will exceed the motolyzer capability.

if you have a multi meter you can measure higher current. If you can figure out how to measure rpm on chassis you can do the same xy plot with some creativity. Telemetry helps.