What's the best 7" & 8" woofers available?

Rbj, the answer to your question depends on the drivers and their crossover points. When using active crossovers, they should be used to bracket the ranges that the drivers are most suited to drive.If you use only the flattest center sections of the response curve for that driver, then you will likely also be free of the typical phase shifting and impedance rise areas of that driver. Use steep enough slopes to have them well rolled off before they reach the "bad zones". The woofer will probably benefit from a Zobel network tuned for the resonant freq(Fs) to deal with the spike around that freq. The tweeter may also have a similar problem, depending on where it is crossed-in. If you try to run any of the drivers in a wide-range manner, you are likely to need it due to your running into the extreme upper and lower freq.s that driver can handle. This is where most phase shift and impedance rises occur with most drivers(there are exceptions).I have found it to be very beneficial in single driver systems also, where the impedance peaks on both ends of the spectrum must be dealt with. Very useful circuits.

Rbj. Your question is a very observant one. If you have followed my threads on the single driver subject, you will know that I have been using a network like the one you describe for my own speakers. Yes, the network does lower the impedance spike at the resonant freq. and allows the amp to control the woofer more comfortably at lower freq.s, thereby extending useable bass response and reducing phase shift. However, trying to use long Xmax with a tight cone suspension merely raises the Fs of the driver and takes you away from your goal of lower bass response. The method I used to get good results was to use a network, and tune a transmission line to a freq. half that of the driver Fs. This allows the T-line to handle the output of the lowest freq.s at a higher SPL than the cone would produce, and with shorter excursions. In my particular system, I achieved a full octave deeper bass before rolloff, with this method. In my network without the capacitor, the high freq.s in a single driver system are also improved similarly, with an extension to the high end gained by reducing the gradual impedance rise that is commensurate with rising frequencies. T-line tuning is a bitch, and has some problems like comb filter effects, but they can be minimized with some work. I think that if more people employed a design like this, you would see more single driver systems being accepted as true full range systems. The nominal impedance value of the driver is slightly lowered so an amp with some flexibility in this area is required. Good question.