Notice how, in this photo, the modeler is carving away a slight amount of
material from the edges of the axle slots, in effect, chamfering the holes?
This will allow the brass axle bushings to "seat" better, and should also
increase the lateral ( side-to-side), motion of the rear axle slightly.
A little side-to-side "play" is OK. But an excessive amount will reduce the predictability of your model in the handling department.
In other words, you want that rear axle to spin easily. You can check this by popping the motor out and spinning the axle by itself. If it's "tight", or doesn't spin easily, try removing a bit of material as shown and test it again. Repeat this procedure, removing a little bit of material at a until it spins freely. You might also want to scrape a tiny bit of material from the surface where the bushing seats on the geared side only, to increase the free-spinning capabilities of the axle.
This model is made mostly from plastic. And plastic is just that: "Plastic"... When something is described as being "plastic" it implies that it's somewhat unstable, moldable, malleable. In the case of injection molded plastics, as the plastic cools, it shrinks. Different plastics have different, fairly predictable "shrink factors". Some are affected greatly by moisture; Others are not. But all are affected by heat.
When we designed the M8d, one of our goals was to keep the tolerances on the tight side. We'd raced enough sloppy models to know that we liked the better control offered by a "tight" assembly.
So when we designed the "Motor Pod" which holds the motor and rear axle assembly, we calculated that it would shrink a given percentage and that would be that. But "Murphy's Law" struck us and the stuff shrank a tad more that we expected. This didn't really present a problem since a few thousandths of an inch, one way or the other, wasn't enough to cause any concern, assembly-wise.
( A bit of digression might be in order here Plastic is not the only thing on this planet that's "not quite stable". The electronic components in your radio or TV are generally what are termed 10% components. Which means they can vary 10% plus or 10% minus from "spec". Odds are the mix of components will be undetectable by the listener. But it could account for why your car radio "picks up" more ( or fewer ) stations than your buddy's. )
(This stretches over into just about all of man's endeavors. From automobile engines to spacecraft, to the frying pan on the stove. Some just seem to work better than others.)
(Another for-instance is, the motors we bought for these models are identical to a zillion others made for the slot car industry. And like the others, these also vary from one to the next. Not a lot mind you, but when everything else "seems to be equal", but one car is still faster or slower than the others, you can chalk it up to that "10%" rule, once again, and call it "the luck of the draw".)
(Which, by the way, was the reason for starting this dissertaion in the first place: The motor mount pods on these cars, being injection molded on hot days and cold days, humid days and not, vary slighly from one another. But the rear axles, being steel, and the wheels being aluminum, are pretty stable stuff, unaffected by humidity. These were assembled on a fixture so the "back-to-back" is relatively constant. So because of the "instability" of the plastic, some fit tighter than others. Once again, chalk it up to the luck of the draw.)