Wood is a thermal insulator, so with care, one might be able to build a thermistor bridge capable of detecting the differential temperature rise between pieces of brass when a source was placed near one brass-piece.
I, too, found that the expectations framed by the title were not met by the article. It's a pretty box, though.
Sidenote: that sounds like even F-22/35 should be visible half a globe away in IR
Even has some links to the model(s) used.
The key ingredients are aspiration and time.
Obviously the ability to skip by basic research and dead ends would cut this down by a ton, but ultimately a lot of the things we’ve created in the past century requires a huge amount of complicated machinery to make, which requires more machines and more labor. Could a person reasonably start from scratch (let’s say skip the mining and just give them coal & metal ores) and make a Geiger tube? Or would it be too difficult?