Draft outline — to expand. The section headings below follow the copy deck (§12, M2). Replace this note and the placeholder lines with the full text.
Ghostwriters and staff: whose voice are we scoring?
Placeholder — a lot of "a leader's words" aren't theirs. What that does to the signal.
Thin and uneven records across eras and candidates
Placeholder — some figures have oceans of coverage, others a puddle.
The reductionism risk: a person is not a number
Placeholder — the danger of flattening a human being into a score, and how we resist it.
Coverage bias we can dampen but never delete
Placeholder — source weighting reduces bias; it does not erase it.
A high score ≠ a good outcome
Placeholder — perception is not performance; the map is not the territory.
How we keep WATT a "directional indicator," not a verdict
Placeholder — the guardrails that keep the number humble.