I've got to question how the driver got into this situation? If he turned the wheel toward the curb and allowed the vehicle to roll into the curb, prior to shutting off the engine, there is no strain on the steering lock. Then, after shutting off the vehicle, a slight movement of the wheel will cause the lock to engage. However, there should be no more stress on the lock that there would otherwise be on level ground.

So, the reverse of this should also be true. The steering wheel should only need a slight pull, one way or the other for the lock to disengage.

However, if the doofus parked against the curb and turned the wheel into the curb with force (because the tire doesn't want to turn into the curb), then some undue stress could be placed on the steering and the lock may not want to disengage.

I can't see a problem occurring if this procedure is effected properly.

Additionally, I always remain concerned with using the transmission to hold the vehicle on a grade. If the grade is sufficient, the shifting mechanism and linkage is not strong enough to remove the pawl (remember the weight of the vehicle is leaning on it). The linkage fractures and the result is a very expensive repair. Additionally, the vehicle is completely disabled right where it sits. It can't even be loaded on a flatbed without extreme difficulty (the drive wheels won't turn).