Language games: [7] I like the notion of game tiles pushed around on a board, as long as they are irregularly shaped & don't always line up perfectly--a jigsaw puzzle with imperfections. Each little intelligible section can be thought of as a "language game" & the whole, finally unsolvable puzzle is the language game of language games. The crucial insight--one that Wittgenstein's logical positivist pals could not accept--was that the puzzle, by definition, contained discontinuous areas. Godel did for mathematics what W did for language. And you have already mentioned Heisenberg's uncertainty principle--when we apply analytical tools to language we change the language under consideration.

The big "move" in 7 is the one that says all the language games together form a language game. Is there a connection here to Russell's Theory of Types?

When two language games come to sit together at a table, palyfulness & caution are in order!