CG and %MAC are the same thing looked at from either the left side perspective, or the right. For the line pilot, they are the same, albeit that the numbers are different.
Then, why ?
Traditionally, the aerodynamicists get to work up the AFM. Aerodynamicists and FT folk are very interested in CG expressed as %MAC as that measure correlates with likely aircraft handling characteristics. If the aerodynamicist is going to run up an AFM, is he/she likely to rework all the CG data from %MAC to distances aft/forward of the datum ? Of course not. The line pilot gets the data as %MAC and that's just tough.
It follows, then, that the line pilot needs to be able to convert %MAC to FS and FS to %MAC. Minor arithmetic sums which are just one extra minor nuisance for the line pilot to do to pass the Regulator's exams and so forth.
Generally, the small aircraft OEM will avoid %MAC in the POH (not sure why) while, as the size goes up, you tend to get it all in %MAC.
Not worth worrying about to any extent .....
Who calculates %MAC?
Anyone who has the relevant data and an interest/need to do so.
How does a pilot know CofG, %MAC of the aircraft they are flying
It is an essential calculation pre-flight so the pilot knows the value pre-flight every time ... or, should know.
Does it change with speed or is it a set value for that particular aircraft weight and configuration?
Varies only with aircraft loading flight to flight and, as fuel burns, during flight.