I assume you are talking about presaccadic remapping and whether RFs jump or spread.
I haven’t read this source fully. It’s on LIP rather than FEF.
There are a lot of possible complications though.
Methods are pretty indirect for remapping studies so there are other ways to interpret results. It might be a spread sometimes and a jump other times depending on the thing being studied (layer, region, cell type, time before the saccade, subthreshold vs. firing, etc.) since it’s different in different layers. On average although with a lot of variation by cell, cells projecting to superior colliculus lose their response to the current visual stimulus as they gain their response to the upcoming post-saccade stimulus, unlike some other layers. It could be a purely attentional effect, so possibly just revealing parts of receptive fields which are less influential normally but still are functionally important.
It might not even make sense to think about this system just in terms of receptive fields, which could make hypotheses about remapping more self-confirming especially because remapping seems very noisy. For example, if you recorded from HTM’s temporal memory, you could argue that all cells in the same minicolumn have exactly the same receptive fields so minicolumns are therefore for robustness to noise (but the noise is actually sequence context). You could find particular activation patterns (which are actually representations of places in sequences) and make arguments about what those activation patterns are for, such as correlations between the activity patterns and pretty much anything (since a lot of things produce different sequences) and claim a result.
That’s not to say that remapping isn’t interesting and useful evidence. It’s a big clue but hard to work with.