This isn’t about the census but on a related area, of protocols for standardised government data collection:
The premise of website is a good idea- design-system.service.gov.uk/get-started/
it seems to offer free user-tested advice and code for government departments to use when they’re building information collection webpages for the public to use in the gov.uk domain. So far so good.
They offer guidance on how to code to collect information around various personal information aspects such as postcode, name, and a couple of protected characteristics including ethnicity, but specifically say they can NOT make recommendations around sex and gender. Now why would that be? 
See this list- look for the heading ‘Patterns’ to see what I mean: design-system.service.gov.uk/community/propose-a-content-change-using-github/#top
And yet this important question of resolving how to ask about sex and gender doesn’t appear in their backlog list of what they’re going to be working on? design-system.service.gov.uk/community/backlog/
Never mind asking about any of the other protected characteristics..
On the bright side, they seek for feedback about successful ways to survey. I would have some feedback about how to make it likely that women can respond (and what puts women off) we have threads on various aspects of surveys and data collection discussing this on FWR.
This government page on sex and gender then says (and remember it’s talking internally to other government officials):
More research is needed on the best way to ask for gender. If you ask users for their gender as part of your service, get in touch to share your user research findings.
Note despite saying there isn’t enough research- which would generally indicate need for caution- here the Government instructions to colleagues are heavily politicised and in a way which we know will compromise data collection:
They say: If you do need to ask, use ‘sex’ when you need biological data ( for example, if you’re providing a medical service ). In all other cases, use ‘ gender ’.
design-system.service.gov.uk/patterns/gender-or-sex/
In all other cases use gender?! Why? What about non medical structural disadvantage to women that gets lost when you no longer identify women as a sex category?
For comparison see their detailed instructions re ethnicity surveying design-system.service.gov.uk/patterns/ethnic-group/#top
Maybe any MNers who know about data collection or coding can propose a change to suggest ideas about how they should ask about this pattern around sex and gender via these links:
design-system.service.gov.uk/community/propose-a-content-change-using-github/
design-system.service.gov.uk/community/propose-a-component-or-pattern/
Note this is via GitHub which requires a name and email address.
You can also write to the officials who run this service here
design-system.service.gov.uk/get-in-touch/
It would be great if government were able to recommend a non-sexist, non-politically biased, non-homophobic, non-undermining-of-safeguarding, reality-based and equality-promoting way of collecting data that worked for everyone, women and trans identifying people. It would go some way to helping to tackle regulatory capture if the stats are not gender-based.
I’d be really interested to see what MNers who know about this area think.