Standardise upon a templating system for transcripts

Currently some transcripts use placeholders like `(user)` and use `.replace("(user)", user)`, whereas others use square brackets and `str#format()`, and I'm not sure why.

Let's try to move to one way to embed templates in strings - either one of the options above, or something like Jinja that'd let us make transcripts even more flexible.

The issue names no files, tests, or entry points. Start by inventorying transcript uses of `(user)`, square-bracket placeholders, `.replace()`, and `str#format()`; done means selecting one templating approach and making transcript formatting consistent.

Written by the indexing model from the issue text.