Boyo Jim–to elaborate on the damage question, the only thing that I can think of that could go awry in the engine is the remote possibility of the timing chain/belt jumping a tooth (or teeth). Typically, if it jumps one or two teeth, the car will run poorly. If it jumps several teeth (which I seriously doubt could happen from one instance of turning backwards), it won’t start, and if it’s an interference engine it could bend the valves. I’ve never actually seen this type of damage, but it’s not impossible.
billy–what I find unbelievable is a car rolling on a very gentle grade such as Boyo Jim described in his 2nd and 3rd posts (“very slight slope,” “in the garage,” “on very slight inclines”). I didn’t specify that condition in my post you cited, as it was continuing discussion from posts where it had been specified.
As for a car that’s in gear rolling from a stop on a steep hill, as the OP suggested, that is certainly possible. It varies with the particular car and the steepness of the hill.
Ringo–in lower gears (higher gear ratios), it’s easier for the engine to turn the wheels than it would be in high gears. In higher gears (lower ratios), it’s easier for the wheels to turn the engine than it would be in low gears.
Here’s a very simplified way to visualize it: in low, it takes 4 engine revolutions to achieve 1 wheel revolution, while in high it takes 1 engine rev to get 1 wheel rev. The engine gets four times as many power pulses to move the car in low than it does in high, making it easier to take off (because the engine is driving the wheels). By the same token when you’re pushing/rolling it, you have four times as many compression strokes per wheel rev to fight against in low than in high, making it harder to roll (because the wheels are driving the engine). (Your correction is right, but it negates your assumption.)