It doesn't work as well with zooms as compared to fixed focal
length lenses.  Diopters work best with fixed focal length
lenses because they have fewer compromises (to be magnified).

The reversed lense is an achromatic diopter.  The power is equal
to 1000/focal_length; which is why a 50mm is the same as a +20
diopter.

A diopter lense in front of a longer focal length produces more
dramatic effects than with a shorter focal length.  Hence
reversing a 50mm lense in front of a 100mm lense makes more
sense than putting a reversed 100mm in front of a 50mm lense.

The best quality regular diopter close up lenses are also
"achromatic" (they have multiple coated elements), and rather
than costing less than $20 each they are likely to be at or more
than 5 times that.  The difference in quality does show up in
the images taken too.  A well corrected 50mm lense is just a
very powerful diopter... with a bit of a useless focusing
mechanism included... :-)

I don't really know. The manual says the VR is useless at 1:1 and I
haven't seen anything to disprove that. The VR is good for butterfly
portraits & 'closeups' but not at extreme macro. I don't have any
external flash so this setup is strictly a lab scene for me with the
camera sitting on a table & subjects propped up in front. My tripod was
useless at this scale with bright window light. I don't know, maybe it
could be hand held in midsummer daylight, I haven't tried.

I didn't try the 105 at infinity, only close to the extreme
magnification. What did amaze me though is that the focus ring was very
useable even at 1:9 magnification with the 2x TC. I had heard you need a
focusing rail but the 105 VR was easy to focus manually at these
preposterous magnifications. And that's the thing that surpised me about
the 105 used as a telephoto lens; how crappy the AF is; always hunting
for gnats in front of the lens when I want to AF on a bird 200 feet away:
<http://www.edgehill.net/1/?SC=go.php&DIR=California/Bay-Area/San-Francisco/neigh
borhoods/2006-10-31-sailing/full-set&PG=2&PIC=8>
I think that one was the 105 with 2x TC, at least I had it out that day
and I recall it was sharper than the 70-200/2.8 at 200.

Huh?  I gave it to my son, so I can "borrow" it anytime I like.  It's too
nice of a lens to sell.

Rita

Well, it should be cheap enough to try, and experiment beats theory any
day. However, from the theory:

Spacing the supplementary lens further from the primary one will do two
things. First, it will almost certainly make any vignetting worse, or
possibly introduce it where there was none before. The effect would be
similar to stacking two filters instead of one on a wide-angle lens.

The second would be to slightly reduce the power of the close-up lens.
The usual formula for the combined power of two lenses

1/f = 1/fa + 1/fb

assumes that the two lenses have zero separation. For a separation d,
this becomes

1/f = 1/fa + 1/fb - d/(fa*fb)

IOW, f gets bigger for more separation, and thus magnification decreases
for a given image distance v (which will be limited by the mount of the
primary lens.

However, the second effect would probably be relatively small, at least
to the point where vignetting renders the whole thing worthless.

BTW, I assume you know that the supplementary lens should be used wide
open, and that aperture should be controlled by the main lens (the one
on the camera body) - but worth mentioning as other readers may not
know. Stopping down the secondary lens will guarantee bad vignetting.

David