dddd

Only needed on one side, since the wheel would need to tilt further than the fork legs would allow in order for the other end of the axle to move out of it's dropout.

Amazingly simple, but can't the thing be inadvertently defeated by being in the wrong rotational position as the QR lever is flipped?

I suppose if the adjusting nut were Loctited to the shaft that the operator wouldn't be able to flip the lever if the device were in the wrong orientation.
But if the nut is backed off then the QR could be flipped with the device in it's "release" position AND with the lever tension perhaps set too loose.

The problem with this device is that it would require the operator to know how to use it, which makes it somewhat redundant to the QR lever.
It would work as intended though if the only setup problem were incorrect tensioning of the QR nut (or if the operator used the QR handle as a wing-nut). That's the only benefit that it offers, sparing the operator from having to determine quantitatively just how tight that the QR lever tension should be set, and not having to turn the "wing nut" (qr lever) tighter than is comfortable on the hand (if used improperly).

The idea of adding complexity to the setup of something that the operator is already doing wrong is almost humorous. But hey, it might work, right?