There could be many reasons on how those rumors come about that we could speculate on, but I feel PLAN might want to validate this clean-sheet design with conventional propulsion before going ahead with nuclear propulsion. We have seen such incremental approach with 001->002->003 so I am not convinced there would be a huge jump when it comes to 004.
I agree with you there, but given how prominent the noise has been in characterizing the construction of a "definitive" CVN and a "possible" additional CV are, I think trying to fit what we see with those rumours makes sense.
Option C: we should interpret this as extra space added ahead of the island on the new design rather than the island shifting aft. The 003 has its stern extended compare to 002 as this was the only way to lengthen the ship without making major modification to the rest of the hull. With a clean-sheet design, this lengthening could be done on the hull forward of the island, and we should see this being reflected on the mock-up.
That's fine and can exist alongside both of the prior positions A) and B). The important thing is that the position of the island on the mockup (keeping in mind the mockup is almost certainly missing a bit of additional stern hull length and deck length) is more aft relative to the new overall hull and deck length and design, than the island on 003 is relative to 003's overall hull length.
Option D: they made enough change to the internal arrangement to allow the island to shift. For example, they could place the boiler rooms/nuclear compartments directly ahead and aft of the rear turbine-compartment, which would allow them to shift the island by 20~30m. However, I see this as very unlikely as it goes against what I call "design heritage". It would be more likely for the internal arrangement to be near-identical to 001/002/003 aside from modification needed to support nuclear propulsion.
That is plausible, and would be an explainer for B).
That said I also consider it unlikely for the same reasons you described.
The key point is that there are multiple reasonable ways in what we are seeing could reflect a desire to test the maximal permutations possible for island placement and island configuration.