ATM Aplanetic gregorian speculation

[Hugues_Laroche@aia.com]

I am still playing with the idea of an aplanetic
gregorian, and, since there was a thread about this
those days, here is an idea that came to me:

The idea would be to make a Coude-like gregorian,
but with a centrally pierced diagonal flat placed at
the focal plane of the primary. The light would go through
the hole in the diagonal (which size will determine the
full-illumination FOV), go on the conventional greg.
secondary, back to the diagonal and out of the OTA.

I made some computations: in order that the
obstruction caused by the hole to match this
of the secondary, the constraints are relatively hard
to satisfy. Typically, the primary and secondary
must be very fast, for example f/2-f/2.5 for the primary.
The total system would be fast, f/4->f/6, with an obstruction
around 20-25%.

An advantage I see regards the baffling: the pierced
diagonal and the secondary can be enclosed as a whole in a
tube that has only the diagonal hole and the exit
as openings.

Of course, the big drawback is the figuring of the ellipsoidal,
very fast primary and secondary. f/4 is already quite
a task... would the experts judge f/2 or f/2.5 reasonably
"figurable" by an amateur?

As always, I guess others already had this idea... if it was
so good, there would be systems like that around?
Opinions welcome!

Thanks, and clear skies
Hugues