I doubt @PutItOut is going to keep this place running for more than 90 days given the $$$ amounts involved.
That pretty much limits us to the following Scenario:
https://www.microsoft.com/en-us/cloud-platform/windows-server-pricing [ $882 ]
https://www.microsoft.com/en-us/store/d/SQL-Server-2016-Standard-Edition/DG7GMGF0DS1D/0002 [ $3200 ]
https://www.cloudflare.com/plans/ [ $20/month ]
HP Proliant DL360e
~$6000 for a decent server (96GB RAM, SSDs, Intel Xeon E5-2450v2)
~$310/month for Network & Colo (20TB/Month)
But realistically, someone would have to put up ~$10k to do that.
A rewrite for Linux is likely to exceed the time left for Voat and (frankly) would only save on the initial $4k licensing costs. If you go down that road, honestly, you are better off taking Reddit's codebase and porting the data, then adding features later.
Similar servers could be rented from places like OVH, Hetzner, Online.NET, etc. for similar prices a month but they aren't really setup for a modern SQL Server kind of config and they'll try to get you to use 2012 or 2008 which is a bit too old at this point.
Long term its ~31% of the $1k / month from here https://voat.co/v/Goats4Voat/1866558 which tbh is probably optimistic. No offense to anyone but I doubt more than 25% follow through ($250/month)
Tbpfh, after look at the options and my estimate of ~90 days before Voat kicks the bucket, taking Reddit's source code as a base and putting it on a linux box from a cheap commodity provider its probably the best/fast/safest exit strategy. Then work on re-including Voat's features on a case-by-case basis as time allows.