You assume contributing via mailing lists is hard, which it is not. In the project the size of a linux kernel it would be failry similar in complexity, and probably way harder to use anything else due to distributed nature of the project. (which I think is unreasonable to think will change)
I contributed various things to the Linux kernel over the years across many subsystems. Each time I only had to prepare a list of recipients and send the patch series to all of them. Often times the patch series involved multiple subsystems.
In the world where one maintainer is using github, the other his private gitlab instance, one is using gerrit, and the third gitea or whatever, and some holdout still accepts only mailing list contributions,... contributing for people like me who don't get paid for most of their work would be so convoluted, that I would not.
While learning how to send a patch series via e-mail, while it takes some time to learn, allows me to contribute to any part of the Linux kernel via a single workflow, that is easy to replicate.
I contributed various things to the Linux kernel over the years across many subsystems. Each time I only had to prepare a list of recipients and send the patch series to all of them. Often times the patch series involved multiple subsystems.
In the world where one maintainer is using github, the other his private gitlab instance, one is using gerrit, and the third gitea or whatever, and some holdout still accepts only mailing list contributions,... contributing for people like me who don't get paid for most of their work would be so convoluted, that I would not.
While learning how to send a patch series via e-mail, while it takes some time to learn, allows me to contribute to any part of the Linux kernel via a single workflow, that is easy to replicate.