Respectfully, this is an example of rigid thinking that isn't productive or accurate.

Apply marginal thinking the way economists do. People's propensity to return to the office falls on a spectrum ranging from 0 ( the people who are already back in their cubicles ) to 9 ( "You'll need a SWAT team to get me back to the office" ).

A 5 on that scale is a marginal worker. Provide an incentive for them to return, like a nice office, and they become a 4 and return to the office. Now give a nice office to a population of 1000 workers who are distributed evenly over the above spectrum. They all move down by 1, and about 1/10th of them return to the office.

Saying "People who like WFH will never want to return to the office" is only true of the 8's and 9's, not of the marginal worker. They are the ones who respond to incentives.