The only problem with the CDFA concept is that you end up with situations like the one 212man describes, where the MAPt is at the beacon in a difficult place to make a landing from if you do happen to get the references. Not only do you not get to the minima but converting to a visual approach to land is made more difficult.
What is wrong with flying an approximate 3 degree approach but adjusting the RoD (but complying with any step-fixes) so that you get to MDA a reasonable distance before the MAPt? That way you can fly the approach to the actual minima and have the opportunity of gaining the required visual references in stable level flight (probably using the alt hold) before the MAPt.
It seems a nonsense to apply FW techniques to RW just for the sake of commonality when there is no real benefit - this procedure is all about removing trim changes and the potential for CFIT from FW approaches - really not applicable to RW.
The idea that all approaches then look the same is really dumbing down what pilots get paid for - can't imagine this will be used in the offshore environment where that extra 50' will make the difference between getting to the rig (and keeping everyone happy) or going home or diverting.
The MAPt is where the calculated obstacle clearance planes guarantee terrain clearance if the MAP is followed - one could argue that if you are still descending as you pass the MAPt (even if you have added 50') you will be below the obstacle planes (albeit probably briefly) as you initiate the go-around - this is not the case if you come in level and go-around from that configuration at the MAPt.