Actually Plasma is generally more popular than GNOME every time surveys are conducted. However we have to keep in mind that the direct consumers of a DE are actually not the end users, but rather the distributors who package and distribute it. There are a number of historical reasons why many distributors ended up picking GNOME over Plasma including accessibility, corporate sponsorship, an easier packaging experience, and the rocky KDE 4 rollout burning a lot of trust. So what you end up with today is many distros shipping GNOME despite pent-up desire for Plasma. It's a great illustration of how you need to keep your direct users happy.

And I think that pent-up desire is being unleashed these days due to various changes in our ecosystem. Plasma is better than ever and version 5 had a much less painful release compared to 4, with us aiming to do even better in Plasma 6. We also see an increasing number of hardware vendors shipping devices with Plasma on it (https://kde.org/hardware/), who had a strong financial incentive to listen to their customers by picking Plasma over GNOME. In addition, KDE's accessibility game is ramping up hugely, and we have more robust corporate sponsorship than we used to with Valve and Blue Systems putting tons of resources into KDE. Finally, GNOME seems to be becoming more hostile to their downstreams, causing them to need to do more of their own development or else migrate to be a fork or skin of Plasma. Interesting developments.

We do aim to improve our design and usability further as much as possible, however, one of the nice things of free software is really this big choice. There are different projects and one size never fits all, if some people find the software written by our friends over GNOME more suited with their needs, that's totally fine.

It would also be interesting hearing on the motivations for this choice tough, as it always help us improving