Mozilla's post on Interop 2026 opens with the number that matters. Interop 2025, the program in which browser makers commit to fixing the same features the same way, started the year with a combined score of 25 and ended it at 95. Firefox alone went from 46 to 99. For 2026 the group took more than 150 proposals and chose 20 focus areas, among them cross-document view transitions, scroll-driven animations, container style queries, and scoped custom element registries.
I think that 25 to 95 is the most important number a front-end architect has seen in a while. For most of the web's history, a feature being in the standard and a feature being usable were different things, separated by years of waiting for the last browser. Interop is the browsers agreeing to close the gap on a schedule, and the 2025 result says the agreement holds.
My view is that this changes the default for a team deciding how to build something. I think the default should now be the platform first, with a library only where the Interop scoreboard says a feature is still short, and a note to revisit when it is not.
Where does your team still reach for a library out of habit, for something the platform now does on its own?
Photo source: https://photos.robertstowe.com/new-mexico