For reasons of both my dayjob (IT at a software company) and personal needs, I have one of everything and multiple of some: Windows, Mac, and Linux. They're all running, all the time, because (as you pointed out) some things run better (or at all) on some systems but not others.
Switching between them has become second-nature, even the Ctrl-C/Ctrl-V on Windows/Linux vs. Cmd-C/Cmd-V on the Mac. The good news is that Copy/Paste works seamlessly between them all, because my hardware is Mac, with a VMware VM of Linux, and Remote Desktop to my Win10 machine at the office. So all three OSes are running on one piece of hardware.
I love it. But at the same time, I hate it -- why the heck can't it be simpler?
I'm 68, and have been working with computers since the early 70's (see my FR profile for details if you're interested).
In some ways, it was much simpler way back then.
Like you, I’ve been using computers since the 1970s as well. My usage was different than your IT background. I was a user person with engineering and science kinds of applications. Let’s give a big shout out for FORTRAN! LOL!
I don’t know if many now appreciate what it was like back in this time in that there were few examples of commercial software available and you had to write your own programs starting with a blank sheet of paper, a pencil and eraser. What programs there were were very expensive. Real real expensive!
Mainframe in college. Switched entirely to handheld HP and TI scientific calculators for the 1st year out of college. Our analytical lab upgraded to a stronger HP desktop late 79 so we grabbed their obsolete HP 9825. The cost of the computer was $10k and for that you got a computer about the size of a typewriter that had a built in keyboard, 64k of memory, a 20 character led display, a cassette tape reader, and a thermal printer using a roll of calculator paper.
Myself and another guy wrote programs to automate calculations for data coming out of our pilot plants. We kept layering on more work to the computer and the bosses kept giving us more $$$. After a year or so, we had $20k invested in that hardware setup via an x-y plotter, 5 1/2 floppy drive (size of a typewriter) and external thermal printer (Size of a typewriter). Our software capability had grown exponentially and gone past individual pilot plant customs to focus on handling generic data and running it through a battery of statistics and plotting results.
Along the line, our corporate IT folks heard about what we were doing and told us we needed to move onto a mainframe. Keep in mind that this was the days of the VT-100 terminal. Corporate had a new and horribly expensive scientific software on the mainframe. We gave it a tryout and said no way, it’s a dog. Well, that caused a bureaucratic ruckus. It ended up that we challenged the mainframe to a trial. A corporate guru sat at a VT-100 and our HP-9825 was at the adjacent desk. We had to enter a data set then run a battery of statistics. At the HP keyboard, I typed in about 50 data sets, ran statistics, printed all the results and had several X-Y plots out before the data had even been typed into the mainframe via VT-100. Eye opening moment. A couple of months later, I was the first person in the corporation with the original IBM-PC. A short time after that, desktop computers were standard equipment corporate wide. Within 2 years or so, the company had several thousand PCs.
A funny thing that trickled down from all this is that I got tagged into a small group of folks that were software testers. My personal contribution was being impatient with BS and so my opinions were valued in the sense that if I could and would use a software, any damn fool could. Hehehe...