however I think the NEC needs to perhaps have a distance away from the tub. Not sure with gfci why this is a major issue anyway.
A GFCI has only limited protection if a load is using a two wire plug and cord, and all plumbing is non-conductive, this has been demonstrated on here and at a few IAEI meetings with a plastic bucket by throwing a hair dryer into a plastic bucket of water with it turned on, and it was plugged into a GFCI, the GFCI didn't trip, but what was also found is that there were voltage gradients around this hair dryer that could have caused a death if a person was in the water with it, the same was not true for appliances with a three wire cord as the GFCI now had a fault return path and tripped.
I for one have never understood why receptacles were allowed close to a tub but not a swimming pool or hot tub, at least a hot tub or pool will have some sort of grounding path to trip a GFCI even if a appliance with a two wire cord fell in, I understand that bathroom design can make it very hard to always avoid putting a receptacle away from a tub, as pocket doors are one of the main problems that can cause it, but even now with most appliances used in bathrooms requiring a three wire cord we still will have many existing appliances that won't and things like radios and small TV's are still being made with two wire cords.
410.10(D) I think is where most inspectors got the idea that we should use the rim of the tub as the boundary as that is the only place it was ever mentioned that defined a tub or shower space, but on the same token 410.10(D) would also be a good guide as to how to write a definition that could apply to 406.8(C)