The story of this merge is not only several years long, but a true success story in OSS in my eyes.
I took on i3-gaps by taking an existing patch and rebasing it to the latest i3 HEAD. From there it became popular and I took on the maintainership, eventually contributing to i3 itself and finally becoming a maintainer there as well.
Whilst originally gaps were considered an "anti feature" for i3, years ago we already decided that we'd accept adding gaps into i3. Clearly the fork was popular, and as someone else pointed out here as well, the Wayland "port" of i3, sway, added gaps from the beginning on with great success.
However, the original gaps patch was focused on being small and easy to maintain. It caused a few issues and had some drawbacks. We made it a condition that porting gaps into i3 would have to resolve these issues. Alas, this could've meant a lot of work that no one took on for the years to follow.
Recently, however, the maintainers of i3 got together (a chance to meet arose randomly). During that meeting we decided that it'd be better to just merge the fork and improve it later. And as it happened, Michael, the author and main maintainer of i3, did all that work during the port as well.
What resulted is the end of almost a decade of i3-gaps, and a much better implementation thereof. I'm incredibly happy to see this happen after all this time, and a big shoutout to Michael here for all that work.
Edit: Hadn't realized Michael was commenting here already. I guess leaving the background and story from my side of things doesn't hurt regardless.