A derived number per solution and scenario, (k - 1) * 100, replaces
'1,068 times as high' with '7 % above' where a reader meets it first.
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
Every measured number in the prose and the tables is read from data/resultater/ by
rapport/lag_resultater.py (modules in rapport/resultater/) and carries file, path and raw
value in rapport/generert/tall.json. Chapter 8 says what is missing: the ranking of the
ownership forms under other weightings and limits is not computed, and the reproduction
is on one machine.
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>