It just seems like things wouldn't fly right, so to speak. If you toss coins at a cup, you could TK them in there, for example, or they could just float away. To do that IRL, it seems like just practice against physics. But physics in dreams isn't always the same.
That's why I thought that.
That's true, Moonbeam and, of course this is one of the big problems with lucid dream practice. But it is just the same as with all other things or tasks you attempt to test in a lucid dream: It depends very much on your motivation, expectations, doubts, focus and patience. it's somthing one could work on.
I have already interviewed some people for my study and some of them have had amazing experiences with practicing motor skills while lucid! Sometimes problems arise, but that does not mean that you can not deal with them or maybe it just works better in the following LD...
Also, you are right: physics in dreams are not always the same as in waking life, but what is more important is that when you perform a certain movement in you dream, the "programme" for that movement is actually carried out in the way as if you were moving in waking life! At least this is what we can assume from dream research and even more, from research on mental simulation, so far. So brain-wise you DO carry out the movement, but your brain stems acutally stops you from acting out the movement physically - a function which is not working for people with REM disorder (they actually DO act out there dreams :( ) This is why you *could* improve...you rehearse motor programmes. Of course, a criticial factor to discuss here is the missing feedback from both the(physical) body and the environment. It can be assumed that the brain manages to simulate this feedback to a certain extent - in that case it would be helpful to have at least some experience with the movement in waking life already - the simulation of the feedback might be more real then.
Regarding the example of coin throwing: I think it does not really matter if the cup in the dream is positioned in the exact same way as in wakefulness. I tihnk it is more important to practice that specific eye-hand coordination while dreaming and thereby improving that neuronal programme.
I think there is plenty still to be shown by research - especially using imaging techniques which are constantly improved so that now participants can actually sleep in the newest fMRI scanners, because they are not as loud as they used to be.