I built up a few of those little boards that arrived yesterday, and commenced testing.
Initial results do look promising: it'll probably solve my problem just fine, and cheap.
When I came up with the initial approach (now, alas, discredited), there didn't appear to be a solution to this problem on the market, nor published. There are commercial solutions to a related, but significantly upmarket, problem, but those are decidedly not cheap, and also require unacceptable changes to Somebody Else's Design.
I can't imagine this being a unique problem - surely other people have the same requirement from time to time! - so maybe the Probulator Mk. II becomes a Product.
For the next month or so, though, I'm being coy about exactly what problem it solves, and how. There'll probably be a companion product, too.