There are several unrealistic issues with these tests which don't reflect the real user's experience:

1. It's done in new installations. Windows tends to slow down over time unless you do some heavy maintenance. Linux doesn't.

2. Most windows installations are bloated at purchase time (with the equipment). You need experience and time to debloat it. If you buy a computer with Linux preinstalled it is not bloated.

3. Most windows applications installations bloat the system even more. Apps tend to install extra stuff and add application autostart entries. Linux applications tend to only install the app.

4. Windows background processes slow down the system unless you do some special configurations to avoid it. Storage trashes for hours with the indexing, rebalancing, etc.

The counterpart is that

1. Other than a few examples, Linux doesn't come preinstalled. And even then, for example, my ThinkPad came with Fedora preinstalled but it was a sloppy installation. The storage wasn't encrypted and secure boot was disabled. I still had to start from scratch.

2. Linux requires less maintenance but what it needs, the user needs to do manually or automate. However, some Desktop environments have started to automate things like indexing and start behaving like windows, slowing the system at boot when is exactly when I want my computer to perform the best. That's why I turn on my computer, right?

I love the fact that these get benchmarked, but I wish that Phoronix made the difference between "lower is better" scores and "higher is better" scores much more apparent.

Maybe put a gradient on the bars so we can see which one is better at a glance.

Or change the "lower is better" layout so the bars are right aligned instead of left aligned.

All operating systems were updated and at their defaults during the benchmarking process.

So it's really a KDE vs GNOME thing?

(ducks and runs)

Seriously, the article doesn't say after 2 pages.

Anyhow, the results are as expected: both Linux distros definitely perform better than Win11, but between them there's little difference.
🤷

I must say, the results are more varied than I expected.