Add the research findings for the strict scalar function layers

Records why there are two authoring traits and not one or three, which trait
members are forced rather than chosen, the strictness-versus-totality finding and
how it lines up with #9033, the problems worth extracting onto develop separately,
and the known gaps the porting probes turned up.

Signed-off-by: "Connor Sanders" <connor@spiraldb.com>

Author: claude