KB1GMX, "Michigan Mighty Mite" http://www.qsl.net/wb5ude/kc6wdk/transmitter.html
N1NQC, I saw your post on QRZ.com too. If you want to learn more about this circuit, it is taken almost directly from one in the 1971 ARRL book, "Understanding Amateur Radio
" (pp91-92) and used copies are only about $5 or maybe you already have one!
The circuit was used in the author's introduction to transistor oscillators. It just follows his introduction to tube-type oscillators as this one is similar to a tuned-plate tuned-grid configuration.
The crystal is used for selective feedback for the transistor from collector to base. The tank is adjusted to allow more or less energy to be fed back through the crystal. Remember, at exact resonance, a parallel resonant circuit acts as an RF open circuit so then there would be no feedback to the transistor and oscillation would stop.
I don't know how you did your coil or measurements, the value seems far off. I input the intruction for the 40M coil into a web solenoid coil calculator and got from 8.5-12.5uH depending on how closely wound it was. This does not seem an excessive value to me. Especially as there needs to be good energy transfer from primary to secondary and also room for the tap.
I don't get where all your worries are coming from? What is wrong with using a polyvaricon? Did you do calculations and saw that its ratings were being exceeded? What advantages can gained by being able so move the location of the primaries and secondaries? Is a super-precise tap location important? Going to Airdux may require a lot more spacing between the two winding and may detrimentaly affect the coupling efficiency since there is no ferrous core.
I think quite a few of these have been built as designed so I think the first step would be to build it as designed. Then do measurements of power, efficiency, signal quality, etc. Then make changes (different coil type, transistor, tap location, total inductance, etc), one at a time, and redo these measurements. This will help you to really analyze and understand better the circuit and the changes you want to try.
The spectral purity part can be really important here because there is no following output filter, and the author talks about tuning the tank for a power vs signal quality compromise. This infers that trying to "max out" the power may lead to harmonics, so I would measure with a SA or at least listen to the rig on its harmonic frequencies on another rig in your shack to check when tuning it up.
73 and best of luck,
I just posted a long submission re this in the QRZ home brew section and just wanted to put the word out here as well. I'll take answers on either site.
ANY Help GREATLY appreciated.
de N1NQC K