And that's presuming that the entities that held the original license still exist today. It's a morass that can be nearly impossible to dig your way out of after 10+ years.
The other problem is open sourcing the Windows code would be also involve open sourcing its crazy huge build system and tooling — it’s not a simple makefile — which is probably another level of impractical.
They were able to get rid of a lot of Windows 1.0/Dos back-compat junk with the release of x64 (due to loss of native 16 bit support). I'm sure there's still some though.
My employer uses 32bit VDIs to run a few legacy applications.
It must be pretty expensive to port/test all of their patches/updates on something few use [0.70% per [0]). They already killed it on their server line (although it was silly on there with the low memory limits). Nvidia has already announced they're ending support[1].
[0] https://www.pcbenchmarks.net/os-marketshare.html
[1] https://www.extremetech.com/computing/267180-nvidia-ends-sup...
I noted in another comment that they have a similar problem with people wanting VB6 open sourced, because so many businesses rely on VB6 apps - some written internally or exclusively for their business.
Windows 10 (from what I remember) almost didn't support VB6, but in the end they included the runtime DLL to allow for the apps to work. I wonder if this is a part of the 32x thing? Seems a likely possibility...
VB6 and the companies relying on it might be an albatross around Microsoft's neck; for how much longer, I'm not sure.
I find it difficult to believe that Microsoft would keep Windows 10 x86 around if usage was really that low.
Backblaze found that about 4.5% of their customers use 32bit Windows in 2017[0].