Unify "protocol objects" language across threat model

[eric-schuh]

Per @TallTed's suggestion in the review:

I believe this is the first instance of available protocols object, which I think is fine, instead of protocols object, which I don't. I think all instances of the latter should become protocol object, as evidenced by several suggestions above.

Looking further down these definitions (circa line 626), I find a few lines about protocols object which doesn't include available in any of the name aliases. It seems to me that the object can be titled with singular protocol even if it contains multiple protocols.

I'll leave that discussion for later.

              exchange URL (<a>D9</a>) to include in the [=protocol object=] (<a>D4</a>) that the coordinator process will create.

Originally posted by @TallTed in #655