Managing windows in a VM with a Linux host on bare metal is long term much more manageable and headache free than dual booting. It’s also a lot easier for Linux to host files to be shared between Linux and Windows than it is to manage a filesystem on bare metal that each alternates access to and to which they can both read and write. Easy sharing of files between systems is going to make the transition a lot less painful. That’s just a lot easier with Linux hosting a VM of windows in my experience. Makes backing up data easier too.
There is plenty of documentation for various options. I have mostly use a mix of samba and NFS to share between various Windows, Linux, Android, and iOS devices for decades. It’s scales well from a single device hosting multiple VMs, to dual booting and accesing shared files on an independent server, to ahomelab with a mix of devices and operating systems, accessing levels, and automatic backups.
All the kids here seem to get really annoyed whenever anyone suggests Ubuntu for “new to Linux” people. My story in particular seems to draw out the trolls, the know-it-alls, and the ricers. I had the same questions as OP 26 years ago, I made the choice you’re recommending (and getting down voted for), I’d do it again, and I have no regrets. Here’s my story anyway in case it resonates with someone.
I picked Ubuntu for my “mostly a server, but sometimes a workstation, sometimes a multimedia PC” before Mint or Arch were even a thing. I knew about and tried Debian, but support for games and hardware at the time wasn’t there for me. Back when we used BitTorrent to literally mostly download Linux ISOs, I was a relatively new Linux user. I’d tried Debian, Slackware, Corel, SUSE, Redhat, etc. Played around distro hopping. But when it came time to build my next machine I landed on Ubuntu LTS mostly because a few important pieces of software I needed to run (paid real money for and needed for university) ONLY came packaged as Deb. Ubuntu turned out to be well documented, well supported, easy to learn, and stable enough that after a decade it was the hardware that failed me, not the operating system. Then, there was the Unity debacle. Then, there were snaps. But, by that time those issues were meaningless to me because I knew I could easily avoid snaps and unity altogether if they bothered me. I never even touched the app store. I guess I stopped caring about the desktop because by that point I was mostly only accessing the CLI remotely or tunneling individual X apps over ssh. When I rebuilt that machine, I considered other options, but ultimately all the choices had mostly insignificant differences except for my familiarity with them. So, I picked Ubuntu LTS again, and it’s been trucking along without getting in my way for nearly another decade.
Arch and those other new distros are interesting. I can see the benefits of that kind of system. But it’s not for everyone. It’s not for me. 99% of users are not going to benefit from bleeding edge software updates. Moreover, there seems to be this widespread misinterpretation that stable and long term release cycles don’t get security updates. These days with snaps, flatpacks, docker, and VMs, running a flashy new bit of bleeding edge software on a long term or stable release cycle distro is easier than it ever has been. It may be slightly difficult for a new user, but it’s still easier than reinstalling and setting up a new distro with a host of undocumented bugs. I can’t even begin to imagine how awful it would be to try to learn about Linux and troubleshoot an issue as a noob in this post-search AI slop wasteland that is the dead Internet.
Anyway, I guess the point I’m getting at is that I chose Ubuntu because it was easy, I chose it again because it continued to be easy, and now that I’ve been using it for a couple decades I’d choose it again because I care more about using my machine than tinkering with my machine. And ultimately, the choice of distro matters a whole lot less when you’re not new to Linux.