Exactly! In fact the software that does that even already exists! I read a news story earlier this year about a team which was using that technique to restore

Exactly! In fact the software that does that even already exists! I read a news story earlier this year about a team which was using that technique to restore old phonograph records that are too fragile to play. They even used it to successfully reconstruct the audio from fragments a record that had already shattered. In the article (I forget now whether it was on NPR or Wired) they mention that the software is flexible enough to follow a linear stream whether the line wraps around a flat record or a cylindrical gramophone; likely it'd work on this medium as well. (I can readily imagine 80 years from now, when they want to reconstruct the audio from an old CD, they'll take a picture of it from 3 feet away with a digital camera and it would be able to resolve the pits on the CD into a data stream.) Many engineers have a tendency to "imprint on" a certain set of technologies and keep tools in their mental toolbox from early in their career. Even most brilliant engineers will do this - technology moves very fast and using the tools you know saves time and effort up front even if new technologies exists that are less time long term. I'd never dream of belittling the hard work they obviously must have done to reconstruct this pallophotophone, but I do notice that in the picture from the article the engineer is not young, and I do think it likely he built a mechanical device only because that's what he knows. If a millennial were doing this project, absolutely he'd try it with software first. But my generation can't be too smug because we're not immune to the effect. Maybe the next generation from now would (to give a random example) first think of growing a piece of biological wetware to do it.