Well, DiMarzio's reply was that due to the space between pickups and positions beneath the strings that it is impossible to get true phase cancelation (got back the obvious without addressing the "why so little loss" or "what is expected loss")...
 
So I guess a freak act of "dumb luck" happened last night. Definitely not going to complain about this one, but pretty much guaranteed I can never rewire this ever again.

Grem
Is that Koa on the DC200?

My G&L Legacy has a push/pull on the volume that does something to the pups in the 3-5 switch settings.

And my DC400 has all the switches like Danny's. And I also gigged with the DC400 and got all the sounds I needed. After gigging with a LP for so many years I often wondered why anyone would want that many switches. But after using the DC400, well I knew then.

I'm so with you there. My DC400 does everything you could ever want from one guitar. I have 3 of them currently and play them more than any of the others that I have in my collection....and I got quite a guitar collection! The pricey Les Pauls and vintage strats just don't move me. Nice to have, cool for some things...they just don't move me like the Carvins do. Hey...whatever it takes to get the job done is what I use. :)
 
This is my latest one:
 

I have been doing some evaluations on this new switch and did a gif of SPAN analyzing a simple run of G(full), Cadd9, D twice (6 chords total). The green is in phase and the red is out of phase - orange is common ground for the two. Setup for this was both pickups both wired in series, and in series with each other, which gives the least signal loss between the two (roughly 2-3dB raw). Clean recording, no FX or DI, straight into my Saffire.
 
This is a nice "look see" for those who have never used a phase switch. (The second Cadd9/D I dug into a little deep on the out of phase track). I normalized these curves, but since I hit the second Cadd9 a little harder, it only brought the out of phase track up by 1.4 dB... that particular hit is pretty much "apples to apples." In the gif below, the in-phase is the LEFT channel, and the out-of-phase is the RIGHT... (can see the variation at the lower right in the video).
 

Danny Danzi
 
This is my latest one:
 

Really nice matching job on that one. Looks great!!

I emailed DiMarzio again and linked the above post for them. This is the reply I received:
 
"My understanding of the theory of constructive interference is that the combination of the signals should result in a combined signal with greater amplitude. That does not appear to be what you are getting, but a guitar is not an ideal test bed, and I assume other factors are present. It may also be worth mentioning that (as per my previous note that two pickups are never completely out of phase) some frequencies may be reinforced rather than diminished, and this may contribute to the results you obtained.
 
I don’t know if what you encountered is an anomaly. At the very least, it’s an interesting exception to the usual performance to be expected from a phase switch.
 
DiMarzio Inc."
 
So I walk away from this realizing I am not qualified to speak to a phase switch's results (the one I have is atypical), but more than happy with the "anomaly" I have created.

Just consider yourself a "Lucky Genius" and have created a new form of "Tone"!
 
They will be writing about you in years to come when old timers get together and speak of that elusive, and hard to mimic "Mettelus Tone!"
 
: )

LOL... the tone part was dumb luck. The rest would require me to track more and screw around less

Just a quick comment to close the loop on this thread. The guy who PLEK'd my guitar dropped a nice piece of advice to anyone interested in these wiring schemes (and are okay to drill more holes), which was "If you ever modify another guitar, you may want to consider a push/push for the phase switch, since you can activate that with the heel of your hand and not have to deal with pulling out a knob."

I asked the question about the push/push toggle. The guitar repair dude said he stopped using push/push switches when one on his guitar broke and the spring shot the top across the stage.

I can see that happening. I am not overly keen on the appearance of them (on a guitar anyway). I also didn't wanted to drill the crap out of the guitar (some of the images on this page made my skin crawl).
 
The push/pull is a bit slower, but because the phase is only "online" when both pickups are online I figured out the most versatility for my setup with the least number of switch changes. Having 7 switches gives 72 possible combinations, but is impossible to really use them all. These pickup are incredibly sensitive, so any cut coil usage got nixed long ago (so dropped it to 20).
 
For my setup, I leave the neck humbucker wired parallel, and the bridge wired series. Leaving the phase switch out-of-phase gives the most tonal dynamics. Then I just use the humbucker selectors to shift around (bridge, neck, bridge/neck-series, bridge/neck-parallel). When is only the bridge or neck, the phase switch is N/A, and the parallel wired neck is a bit more mellow. The parallel neck also takes the "sting" out of the out-of-phase in the  bridge/neck-series (but good for leads), and inserts a slight timbre shift with bridge/neck-parallel. This knocks the switch configuration down to only 4 positions without gain changes (although I do push/pull that pot since I am still getting used to it).