There's no point in converting the existing call sites to use
deterministic closeOverKinds if they never linearize the set.

Test Plan: it compiles, this is basically just documentation

Reviewers: simonpj, goldfire, simonmar, austin, bgamari

Reviewed By: bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D2191

GHC Trac Issues: #4012