Windows 10 introduced support for upgrades when user profiles are NOT in the system C: drive.
It's also possible to mount a partition on an empty directory, though that hides the disk usage in "This PC" it'd have the advantage that there's no drive letter or repetitive registry entries per user like you would have moving folders like Documents & Downloads individually, also it works for MS logins where visitors are invited onto the PC.
Is there some Windows gotcha to simply mounting a whole dedicated partition with all the C:\users data stored in it, like you'ld do in Linux/UNIX?
Reasons to want to do this, include using 2 disks instead of one, so programs can be read on one device, whilst a different one is fetching/saving user data. It'd make backups simpler as C:\users is a data partition, whereas most of C: is about system recovery handled well by images & restore points. Finally a bad reason is that UNIX version 6 supported this in the 70's, so in 2019 I feel like an OS ought support this.
Surprised not to find hits for overlaying a directory in Superuser or Google, just a load of recommendations about clumsy moving of directories manually per user.
I can experiment on a sacrificial machine but perhaps others have been brave enough to try this out.