Bumpy,
My apologies for the late reply. Initially, I was going to recommend another circle as my focus is less on the design side, and more on the flavor of the ingredients. However, it appears that you more or less received what you were after. Even though, there is a steeper curve...
While most of the information here is vague, some slightly incorrect, and a tad misleading, it does represent what you are up against. One thing that was touched on in part, but not completely fleshed out is the phase rotation. None of those reactive components completely block the signal. The tweeter does not stop playing below the upper XO point, nor does the mid stop above it. So we have two drivers playing the same information at a fixed distance from each other. That means they will create both peaks (coupling), and dips (cancellation). These can be minimized by rotating phase. Inductors, and capacitors rotate phase in opposite directions. They rotate almost 90 degrees as they move in and out of full reactance. So, the cap on the tweeter has to work with the inductor on the mid. The cap on the mid has to work with the inductor on the woofer. Electrically, the cap on the woofer even finds itself in series with the mid an octave or two above and below the lower XO point.
The best advice I can give at this point is this. During your listening tests, stick with the same brand/series of cap and coil. It is OK to use a different cap on the tweeter from what you are using on the woofer, but keep that tweeter cap the same brand and type as you change values. Do not bypass caps, but stacking similar values to achieve a custom value is OK. Burn new parts in before you do your listening test. When you have arrived at your final values (and/or topology), you will be ready to select the specific parts that possess strengths for your subjective and unique requirements.