You do have to be careful as floating point or Dintegers are two consecutive 16 bit registers, wherever possible I tend to make 32 bit (float or doubles) on the evens boundary in other words D400, D402 etc, not that it matters it's just when programming on the fly it is easy to forget but a simple xref will usually find any not used just as a matter of interest of learning any instruction that is either float or double integer the function will either have "E" or "D" in front of it, i.e. MOV = 16 bit, DMOV = 32 bit double integer & EMOV is a float.
If using SM400 and it is only needed on the first scan that needs to populate a register i.e. a starting point that may change then you do need to use the pulse of that move function, if using SM401 then there is no need to use the pulse as it is only on for the first scan, if that register never changes then there is no need to move a fixed value into a register just use "K" (constant) in the logic. I assume that it must change somewhere as I find it a little perplexing if it is not being changed elsewhere, one other possibility is that it is a pre-set or default value & can be changed on the HMI, from what I have seen of the code so far the original programmer may have been under pressure & did a number of changes on the fly, believe me been there done that, looked back later & thought why did I do it that way.