[William Cloud Hicklin]

Well, it was a fixed data-point that Frodo observed the full moon setting over the Forbidden Pool, and at least as early as Time-scheme S this was the full moon of February 6 (RW Feb 1). With S2 and the one-month delay this became March 8 (RW March 3). This is a basic "fact" that never changed, throughout all of his permutations.

What did get changed around a lot was Pippin's observation of the same moon, since in the original conception he was already in Minas Tirith. However, it was explicit from the first that this was the full moon, and the same one that Frodo saw. The revised Pippin-on-Shadowfax version required some juggling, but Tolkien got there, and this (second) timeline was what made it into V/i. In the third timeline, as laid out unambiguously in S3 the full moon occurred absolutely on the 8th just after midnight, not on the 9th (by which time Frodo had left Henneth Annun on his way to the Crossroads)

The final or S3 timeline, remember, was developed after the book had been completed in draft. Imagine for a moment the difficulty of Tolkien looking back over a million-word book, written over the course of more than a decade, and without an index much less computer search having to remember all the places that dates or moons were referenced and had to be changed! An impossible task, and he didn't get them all. So therefore we have Aragorn's "three days" across Rohan, for example. He clearly did not go back and revise V/i even though technically he ought to have- for instance "so the night was not yet old" is plainly wrong; there is no question that in the final timeline the incident occurs somewhere between midnight and 5 a.m. not in the evening. And, again, this was the same moon on the same night as Frodo's, an essential conception in every phase.

The relation of the muster of Rohan to the moon was always a mess; do keep in mind that the key passage in "The Voice of Saruman" was written all the way back in 1942!