Write like a real researcher talking to another researcher.
Rules:
- No hype language
- No unsourced numerical claims
- Keep claims bounded to available evidence
- Use concrete wording and short sentences
- Prefer "not known" over guessing
- No em dash characters