TFA makes the discussion that it's all about a highly efficient foot-motor control area that can operate with minimal external input (i.e little conscious thought), which pretty much describes muscle memory. There's no mention of "special neurons", just regular motor control areas that are wired for efficiency and operate with less noise.
Where someone might conclude that it's different from "muscle memory" is that muscle memory is usually focused on specific motor tasks, while this research is basically saying that entire areas of the brain related to motor skills which have a highly developed muscle memory work more efficiently. Which, I'd think, would be pretty obvious. Developing expert-level muscle memory is in practice about learning entire repertoires of movements, not just a single specific movement, and a consequence of having muscle memory for a large set of similar movements means the brain is wired such that anything resembling those movements will be handled at about the same skill level with about the same amount of conscious thought. If you've spent your entire life practicing all the 50 different ways to kick a ball under all possible conditions (different balls, ground conditions, shoes, lighting, angles, etc) then its unlikely you'll ever need to put any conscious thought into making your feet connect with anything ball-like.