Hello folks - I can't speak to why, how and who prefers what balance in their setup. I am glad that Robert preferred the 2.7 as designed.
I can speak to what goes into the stock balance. Two+ years of Thiel home team, knowing very well Jim's goals, design approaches and musical outcomes were invested in the speaker as-is. As background, note that those resistor values under question are complemented by the resistances of the shunt-to-common resonant filters. Changing the feed resistance should be balanced by addressing those shunt values. Note also that those balanced circuits form a mirror of the Thiele/Small parameters of the drivers being filtered. Changing the net values will unbalance the filter with its driver load characteristics. These imbalances are heard as what one of my listening partners has called 'slow jitter' - a kind of instability in the music.
Note also the make-up of those midrange and tweeter resistances. Note that there is a coil in series with the resistors. That small-gauge winding wire is used as a trimming element in that circuit. Ever wonder why Thiel coils are gauged? (14, 16, 18, 20, 22.) Not an accident and not cost-saving. Now, look at the resistors in question. Notice that they are not single-value resistors, but a pair to trim to some non-standard value. The midrange has a 20awg coil (with its resistance) plus a pair @ 8.5 and 2 ohm for a net of 1.62 ohms. Similarly for the tweeter. Probably obvious that I'm highlighting that these values are tweaked, and tweaked, and re-tweaked after the finalized product logs some use-history. In this case, Hometeam Thiel worked with Warkwyn/Canada and the Canadian Research Laboratory to trim to neutrality. All this is to say that someone's preference for some other tweak is their business, but getting it right is far beyond the abilities of most.
My personal experience of the 2.7 includes evaluating the final prototypes on their return from Canada - in direct comparison with their reference CS3.7 in Thiel's music room using (among others) Jennifer Warnes singing Leonard Cohen's 'The Ballad of the Runaway Horse' in duet with Rob Wassermann on bass in pure mic-feed presentation. The presentation of the 3.7 and 2.7 was extremely similar / undistinguishable in tonal balance. Relative shortcomings of the 2.7 were in the areas of textural and harmonic detail which I attribute to big electrolytic caps which were out of bounds in Jim's approach.
Let's look at why those caps are there - it's not just cost, it's that the fundamental design requires them and this fundamental design would not have been considered by Jim. But, Thiel Audio needed to prove its ability to produce a follow-on product after Jim's death, and this model 2 was their best shot. I say they pulled it off. But looking with a long lens, the niche of the model 2 has always been that smaller rooms and lower SPL allow a smaller midrange and woofer, both allowing intrinsically lower distortion mechanisms. The 2s behave better with their easier demands. But this product borrows the 3.7 midrange which was designed to cross an octave lower than the 2.7 into the model 3's larger 10" woofer. The 2.4 (and 2.7) woofer was optimized for rolling off nicely an octave higher than the model 3. So, shoe-horning that midrange into a box it was not designed to fit requires some brute force. BTW, the woofer circuit doesn't have the same problem because it is the 2.4 woofer designed to cross where the 2.7 crosses.
Let's take a little trip. What midrange should the CS2.7 (2.8?) have. It would be the coax Jim was developing for the CS7.3 which would have smaller diaphragms than the 3.7 version. The model 7 low-cut is much higher than the model 3 and would work well with the mythical 2.7/8. Note also, the tweeter diaphragm could be and might have been smaller for considerably greater integrity, and with higher frequency breakup. In Thiel-land the tweeter design must accommodate high excursions in 3-digit frequencies due to the 6dB net rolloff. Jim wanted a 2.4-like mechanical crossover for that coax. New thought could have produced a next-generation passively coupled mid-high coax eliminating the electronic XO circuit between the midrange and tweeter. Now, pair that with an improved 8" woofer. Can you smell magic?
