The tester in the article says he hasn't had this happen on his daily use machine, and that it seems to only happen under very strenuous workloads. Most people should be fine but there's still a risk.

Anecdotally, I'm using Windows, haven't had any issues. I also tried running https://github.com/suaefar/ryzen-test in a VM for only about an hour at 100% load and had no issues.

I would wait. I bought a 1800x Ryzen and had some stability problems under Debian Buster which disappeared after a BIOS update, but I'm not compiling large projects continuously nor it has been up for a lot of time (maybe 3 days uptime? Got the new computer two weeks ago.. when it gets to 100+ days uptime I will be happy)

Under Windows it worked flawlessly even with the old BIOS.

personally, I would say yes.

I've been happily using my 1700x on my personal machine for months without much issue (except when as noted above, the issue was induced with very very high CPU load, manually and in a controlled manner).

Unless you do very CPU heavy work that lasts several minutes or longer (like kernel compiling, chromium compiling, autotools, starting electron apps), especially when you run gentoo (most likely), it should not be an issue.

I do a lot of dev work myself, so I do compile often but a short compiler run hasn't ever induced this bug and longer linux kernel builds suffered more from my inadequate cooling solutions rather than the CPU having bugs.

I wouldn't go for hardware with known flaws like that. You never know what you are vulnerable to. Who knows, maybe someone finds out how you can abuse this flaw with JavaScript or something. It seems to me like AMD bit off more than they can chew.