The question 'What is your religion?' seems like a leading question, since it treats the answer 'None' as being part of the same paradigm as is, say, 'Christianity': and that is clearly wrong since 'None' isn't a choice of religion.
Perhaps using a little more paper is justified so as to avoid suspect data, which would be the case if religious affiliation is over-estimated.
But you are missing the point which is that the ONS state that the humanist formulations and definitions are different to those which the ONS are using.
So for instance Humanists are not interested in whether people attend church or see themselves as cultural christians but whether they truly believe in whatever religion. That is of interest to evangelicals and humanists but it is not what the ONS regard as defining affiliation. The Humanists have an agenda and it doesn't match with the ONS.
As I said, the Humanist definitions are too belief focussed for what the ONS need.
The ONS acknowledge imperfection but classify other formulations less optimal.
As to the performance of people on here........strong on humanist polemic and the ideal of no religion, weak on challenging the ONS formulations and their justifications.