I don't think the person you're arguing with did make it sound easy to do what it seems the paper authors did: which appears to be having the idea to apply this

I don't think the person you're arguing with did make it sound easy to do what it seems the paper authors did: which appears to be having the idea to apply this sort of analysis. As with so much development there is a massive gap between 'I can do that now, someone showed it worked' and 'I could have come up with that before anyone showed it worked'. Leading edge science becomes demonstrations that undergrads do after a time. Plenty of fields seem to progress when someone realises a simple idea can be applied in new way. 6-months to learn matrix arithmetic algos sounds like that person just wasn't cut out for that task. Best option here seems like encourage the person to try applying their domain knowledge from CV to the field at hand and ask them to report back how they get on. If they've applied Hough transforms before, perhaps, they might be able to refine the algos here, for example. tl;dr I think you're being a bit precious about it all.