Checkpoint Not Complete: Every Commit Waits, and the SQL Is Fine

Why an Oracle database freezes every writer for seconds at a time while the SQL is fine: online redo logs too small for the write rate, the log writer lapping the checkpoint, and the alert log's 'Checkpoint not complete'. How to spot it, how to resize redo logs online, and a lab where the same update takes 14.1 seconds on two 20 MB redo logs (11.9 of them waiting) and 1.9 seconds after the fix.