I think this is interesting-- are you insinuating that y rotations don't have regrips? But also, yes, a wide D move is an actual move and has the potential for a lock up while a rotation does not have the potential for a lock up. Though I think that because you don't maintain a hold onto the cube while rotating that it could potentially have a higher chance of being dropped-- but I don't think it occurs enough to warrant not using rotations.A wide d move usually involves a regrip. It has the potential to lockup or ruin the flow of a solve.
A <y> rotation only affects efficiency in semantics (and even then, a <y> rotation could contribute to movecount depending on the metric). Pragmatically, your efficiency in HTM literally doesn't matter. You still did the action, whether or not it's recorded in HTM. For this reason, as a way to more accurately describe how many moves a solve took, some other metric's have been developed, namely (at least relevant to this situation) ETM and ATM.But if you think about it efficiency wise a y Axis rotation can be the main factor.
I have the luxury of using the ZZ method, so for the vast majority of solves I don't need to deal with y rotations. The only difference between a wide D move and a y rotation is whether or not you hold onto the U layer while rotating. As such, I think a wide D move is only useful when it also sets up your following F2L pair, if even then-- y U isn't a bad combo either, and in some cases I'm sure it would get executed like a wide D move.
I think this is interesting-- are you insinuating that y rotations don't have regrips? But also, yes, a wide D move is an actual move and has the potential for a lock up while a rotation does not have the potential for a lock up. Though I think that because you don't maintain a hold onto the cube while rotating that it could potentially have a higher chance of being dropped-- but I don't think it occurs enough to warrant not using rotations.
A <y> rotation only affects efficiency in semantics (and even then, a <y> rotation could contribute to movecount depending on the metric). Pragmatically, your efficiency in HTM literally doesn't matter. You still did the action, whether or not it's recorded in HTM. For this reason, as a way to more accurately describe how many moves a solve took, some other metric's have been developed, namely (at least relevant to this situation) ETM and ATM.
- If you're using HTM (any outerblock turn done on a cube), both <d> and <y U> are 1 move.
- In ETM (everything is recorded as performed), <d> is 1 and <y U> is 2 moves.
- In ATM (any number of moves in the same axis count as one move) however, both <d> and <y U> are 1 move.
tl;dr
only use wide d if you need to cancel into the U move anyway, otherwise just rotate