If it's really just the angle, then I should be
able to get tropical levels of sunlight by rigging up a suitably
angled mirror to reflect the light. eg. at the poles, a mirror of
1m*1.41m at 45 degrees would catch 1 sq m of light and reflect it to
the ground as if it had come from directly above. Maybe this could be
designed into greenhouse roofs so that they had one side of the roof
angled with mirrors instead of glass to give the effect of direct
sunlight from above?

No, because a shallower angle means the light has a longer route through
the atmosphere, so less of it arrives at the earth's surface. It's not
the angle of incidence in itself that matters, it's the effect that that
angle has in determining how much light gets here.
--

OK, imagine an torch pointing straight down onto your work surface. (Or do
it with a real one!) Now imagine it tilted at an angle - it's obvious that
the same amount of light is spread out over a much greater area of the
surface. Likewise there is less light energy per unit of surface area when
the sun's lower in the sky.

As Kay points out, there's also the effect that when the sun is lower in the
sky, the light has to travel through more atmosphere is and gets dissipated
by this. So you can't use an angled mirror at sunset to get the effect of
noon at the equator!