Consider how hard it would be to 'unlock' those joints if it really did lock into place in a mechanical / robotic way, though. That doesn't really make biological sense to have a mechanism like that, particularly in such a vital body part. Adding friction does make sense, as it may take a little more energy to move those parts when 'locked' into place, but in the grand scheme, that extra energy to move when in the 'locked' position is probably negligible overall. Especially considering that it likely doesn't happen often that they move around when in that position, and that the mechanism also relieves some of the energy used to hold those joints in place when they aren't moving around, so it probably balances out. BUT, this is just extrapolation on my part from what I know from physics and biology, so take it for what you will.
Friction is added through the 'rough patches' along the tendons. It's like (in an extreme case) rubbing together two pieces of sandpaper compared to rubbing together two pieces of regular paper. The sandpaper has a lot more friction and it's harder to rub those pieces together than it is with regular paper, which is smooth and slides relatively easily across another sheet of paper. I would assume by the fact that it develops fetally and before mechanical motion could play a role in their development that these rough patches are always in the same general location along the tendon and of the same 'roughness'.
From my experience, their feet are flat when on the nest, both when brooding and when simply there to lay an egg. As I said, those rough patches exist, but they don't override the bird's control of her muscles and ability to move her feet.