[Argus25]

Yes that's right, if the specimen coil doesn't oscillate just reverse the coil connections in the collector circuit.

Practically any RF signal transistor will work in the local oscillator, even the early germanium RF types with high collector to base feedback capacitances.

In the case of the IF coils of course, at least using modern silicon transistors (or good germaniums with a low feedback capacitance AF124 etc) no neutralization is required.

So if it happened that the phase of the winding of the secondary of the 455kHz IF transformers, happened to be reversed compared to another transformer specimen, nobody would know and they would work just fine.

The winding phase of the IF coils is only important if it is a vintage transistor set, say with something like OC45's, where there is a feedback or neutralisation capacitor which takes it signal from the base of the next stage or IF transformer secondary.

By the way, the circuit in post 22 has a poor arrangement for the bias for the transistor driving the primary of the audio driver transformer. The collector current would vary widely depending on the transistor's individual hfe and be unstable to temperature, It needs an emitter resistor to stabilize it , bypassed at audio frequencies with a capacitor and a base to ground resistor as well. It is an example of very poor design.