Hi Charly,
The main problem l have with the liminescence dating, is the small sampling done in the Osireion. Only two samples were taken of the Osireion RHO-1389 (sandstone) and RHO 139 (granite), The sandstone was given a wide range of 1300 +-570 BC, the granite 1980 +-110 BC, and even doses had to be subtracted from these samples to allow for burial in sand. The granite sample from the image in the report,was taken from near the top of the north west granite pillar, the sandstone sample, just says outer wall, though the authors say "The samples were mostly taken from the lower blocks of the monuments or from part of the wall with no indication of disturbance, and with care to avoid light exposure."
I dont know were this sandstone sample was taken, but if an outside wall, and giving the subterranean nature of the Osireion, it would be far to say that it would be burial in sand for a considerable time whereas the granite pillar, may have been exposed to sunlight for a considerable time when the stone robbers removed the roofing stones, and then buried in sand. The variables in this dating method are complex and l am certainly no expert in it; but l feel these two samples are hardly conclusive and a much larger range of samples needs to be taken in the Osireion, and in context with the building and destruction sequence of the structure.
There is nothing in the Osireion to suggest different phases or adapted earlier constructions. To me it comes across as a coherent well executed plan, and as Frankfort states, one can hardly separate the granite construction from the sandstone, afterall the granite is laying on walls that hold Seti's dovetails.