> There seems to be little proof that bees of a single given type fly faster
or slower just according to their size.

Nor is there proof given here that they fly at the same speed or carry the
same payload.  In fact, it seems quite obvious that there must be a
difference, but we don't know what it is.

> if smaller is truly better and faster, then why, after 50 million years or
so of selection, are bees not now the size of gnats and fly at mach 2?

There are limits to any design and as size increases or decreases, these
limiting factors affect the competitiveness of the critter. A good example
is the obvious fact that there is a limit on how large an insect can be and
still walk on water, using surface tension.  Moreover, an insect must
function in an environment, and match the needs of the flowers etc. as well
as function in the presence of predators, etc.  Air density and other
properties vary with altitude and humidity, and over time.  Airplanes that
are designed for high altitude flight are not optimised for low altitudes
and vice versa. Increases in speed have their own drawbacks, and there is
likely a most efficient speed range for a bee.   Air temperature, heat loss
to radiation, convection and transpiration, solar radiation and humidity all
have effects that are exagerated as size diminishes.  The properties of the
air and the bees' environment determine what scale is optimal for the honey
bee.   All these interacting factors determine the optimal size range for
honey bees, and it is a range, as we have seen here.  I realise that I have
not answered the question, but think I have bracketed it quite nicely.

> Perhaps thoracic density is also involved here and may be an indicator of
a genetic type (or supersisters within a colony) rather than strictly being
a size related criterion.

A study Dee showed me seemed to show that bees of the same queen raised in
larger cells were less dense than their sisters raised in smaller cells.

allen
http://www.internode.net/honeybee/diary/