I found this from a discussion elsewhere when the Census results were released a couple of years ago: it appears that up to 118,000 of that count could be false positives (I'm sure it's substantially less than that, but it just shows what a stupid question it was).
"262k answered "No" to the question of whether their sex and GI were the same.
Of those, 118k gave no further information, and AFAIK no link to other census information (that might give information on birth sex) has been made - or at least published.
Of the remaining 144k who did give further information, 48k said TM, 48k said TW, and the remaining 48k said either NB or another neogender.
So we don't know what the split is of the 118k. It's likely that a proportion of them didn't understand the question as per newspaper reports, but not all. All I think we can say is that of those prepared to give further information, it's a tidy 3 way split between TM, TW and other including NB."