Reconsider the browser specific failures graph

[gsnedders]

The WPT Core Team decided last year to defer RFC 122 to the Interop Team, and there was an expectation this would be dealt with after Interop 2023 launched. Thus it's now worthwhile looking at this; I'd suggest sticking to comments in the RFC repo rather than here, but adding this here partly as an Agenda+ item.

[foolip]

We never merged https://github.com/web-platform-tests/rfcs/pull/120 which is where we wrote down that the interop team would take BSF over:

The Interop team will take ownership of the existing Browser Specific Failures metric, and manage it according to the rules in its charter.

In the last discussion it wasn't clear that everyone has the same expectations for how the interop team's consensus decision making would work here.

I think the straightforward interpretation of the charter is that if we don't have consensus then there is no decision and no change. This would apply to Interop 2021/2022/2023 dashboards and BSF.

I would like to improve BSF and somehow avoid the inflated weight of some features, but not in a context where failure to agree means deleting it entirely.

If we can agree to that, then we should make the interop team charter and take the discussion in the team.

[foolip]

Additionally I would suggest:

• A new front page for wpt.fyi, explaining what it is and linking to common views, like the results tables
• Putting BSF in its own view, not on top of the results tables

The front page change requires an RFC though.