KIrk,
I have a look at your drawing.
I am glad that this type of spectrometer approved by our community. I
will search if this is a new type of spectrometer.
More details however need to be precise.
1). First of all, as far as the building of trajectory, one needs to
remember, that magnetic field extends
from the magnet edge to ~half of its aperture, if rectangular model is
used. This is a rule of thumb.
So that means, that the center of the bend is sitting in our case at the
cross point of edges of coils, not
the edges of iron. So that is why the equilibrium bending radius is
bigger, that in yours drawing.
I attached this comment to you drawing:  kirk2.dwg.
Shape of the pole is also needs to build taking into account this factor
(shown). Better to keep it as in my drawing.
2) You also modified the shape of plate. This brakes cylindrical
symmetry is the bending area, however.
It might be small, but I think that extra 20 kg of iron which makes the
field less homogenous is a problematic.
Right now you made the chamber thicker and the chamber itself may serve
as a spacer.
3) The distance of solenoidal lens from the target is 46 mm, I corrected
this in your drawing.
The distance between lens's yoke and the coil of magnet is 18 mm, also
shown. This distance required
for reduction of interference between magnet and lens fields.
4) The movers I hope has the proper size, used at SLAC, otherwise it is
not clear is there any interference between
the flag mover at the center region and the magnetized iron of
DESY/Efremov magnet.
5) We need to remember that flanges in the region of lens must allow
installation of the coil. So probably
wilson type connection to the target box is the possibly easiest here.
More widely, if we do not need baked vacuum chamber
we can switch to O-ring vacuum sealing. This may bring a Stainless
Steel- free chamber.
6) I think that a possibility to have a box in the region of second
target is useful thing. Anyway it is useful to measure
positron signal   as electrical one at the flag (insulated). Also what
we decided about the stray field, generated by magnetizer?
Target located at the maximum of these fields-peculiarity of design.
This might be a box similat to the the one with primary target.
7) Do not forget about chamfers at the edge of pole. Basically in my
design the location of these chamfers defined
relative position of the poles, so in calculation the magnets shifted a
little in longitudinal direction.
So the iron pole has angle between the gamma line an in transverse
direction not 90 degrees, but 86. But this is matter of
exact location of chamfer at the pole. The chamfers at one magnet must
go at the same distance in another, sot his may bring
slight shift of magnets. You will se this, when you will come to the
drawing with chamfers.
Anyway these chamfers have mirror symmetry with respect to to the plane
running through the middle of spectrometer.
(~Where the movers located)
8) I concluded that as our DESY colleagues will build magnetizer of
theirs own design.
I feel free in this case to make publication about my optimized
magnetizer.
This will help people in a future. At the left side of the drawing I
made a 3D comparison between DESY/Efremof and optimized
magnetizer.
 That is all for comments for the moment.
    Sincerely
                Alexander Mikhailichenko.