How `Iterate::next` spells one signature for both item kinds

Decided

D-0021, settled by the language’s author: part of spec §18. Older text cites it as 18.21.

The question is void — there is no next. Iterate is valid / get / advance, so termination is not encoded in the item and Item is never wrapped (§8.5). This reverses the decided half of the old 18.21 (“a pointer iterator returns ^T, a value iterator returns Item?”): that split existed only to dodge (^T)?, and with the item unwrapped it buys nothing. Item may now be &T, which Item? could never express, so §10.4’s for e in &entities yields a &Entity as it always claimed