Well, that's not the "original" 5150 floppy controller. The original one has a number of metal-can IBM hybrid circuits on it. The one you have is a later version of the controller.
Sure! That has already been clarified. This PC has been manufactured in 1984, so it has a newer revision of the controller.
It's "original" in the sense that it has been assembled by IBM at production time (as opposed to being assembled/replaced by a third party later on).
Btw, playing with the old lady I found there's still a residual issue.
One of the FDD (I chose to be A) is totally OK, while the other (B) works fine until the header stands on track 1 (e.g. issuing a DIR command), but it fails if the header has to seek any file. In such a case the FDD header bounces up and down a few times, and after a while an infamous "Data error reading drive B" is issued.
If you have 1'40" to waste, here's what happens: https://www.youtube.com/watch?v=OBXOZCXOVZ4
You will first see the elegant drive A: header nicely and flawlessly performing its job, and then the disgraced drive B: header miserably failing.