We don't generally include spec tables or BCD in guide pages, only in reference pages.

That's what I thought too — hence why I didn't include it originally.

However, I noticed for Permissions Policy we do include these.

I think that while a technology is "experimental" (like Permissions Policy and Compression Dictionary Transport are) that's a pretty good idea to mark it as such and include the full things.

The rest of the HTTP guides really are for core concepts that are supported by all browsers, so including BCD is not really necessary — though specs still seem like good ideas to me in some case. The only other similar case (that probably should be marked as experimental and treated similarly) is HTTP Client Hints, but will leave that for now and stick to Compression Dictionary Transport for this PR.

WDYT?

@wbamberg I looked at the bits that were associated with other work I am doing, but not the rest since you reviewed that.

This bit is the only outstanding discussion and is waiting on you.

FWIW We are not consistent on this. I think the rule is understood, but isn't followed because it isn't that helpful to readers.

I've tried to ask for more feedback from other maintainers on this but not got much in the way of definitive responses.

But I'm tired of arguing about consistency and structure, and I don't think there's anything to say that hasn't already been said, so I'm not going to block the PR on this.

Fair enough. I tend to accept things if I think they are good or not actively harmful in cases like this. Not worth banging your head against the wall for.