The oscillator now works nicely. Early attempts, it turned out, were running on totally the wrong frequency, that's now fixed (with a different circuit) and the waveform looks nice and puts about 1V peak to peak into 50 ohms which I think is 2.5mW.
The next step is to raise this level up to a few watts so that I can be heard at least by VK2BVS who lives near by.
In a burst of foolish enthusiasm, I added a gain stage followed by a class C driver and an IRF510 final, rather messy but the box will be good for the next iteration.
Alas the final oscillates magnificently, presumably my construction is far too spread out. I'm not deterred by this and will have another go this time using a design from Drew Diamond rather than random circuit snippets found on the internet.
I feel I'm making real progress, despite no outright success, my debugging skills are improving with experience.