DoN. Nichols writes:

Then select
a pinhole based on the tradeoff between perceived spot size and
brightness (you need sufficient brightness so you don't lose the spot.

I don't think you get the point.

Making a pinhole shadow of a big blob doesn't make the blob any more
accurate. The blob moves around and the pinhole stays illuminated.

The whole approach just defies elementary principles. You can't improve on
the beam divergence angle, and that is many thousandths of an inch at the
spindle distance scale.