Another game that did a similar thing (that I'm a HUGE fan of) is Spintronics [0].
It goes a whole step further from just creating analog circuits digitally and lets you build them both online and in board game form. I backed it on kickstarter and was amazed at how well the game turned out. Haven't finished it yet due to life getting in the way, but if you're looking for inspiration, their circuit simulator is available online for free.
As a fan of Zach-likes, I'd definitely be interested and play what you're cooking up :)
I find the physical Zachtonic games to be the most compelling, Opus Magnum especially, because you're essentially learning a brand new way to compute using space and time as your algebra rather than logic gates. I'd love to find something more akin to that.
The key to a fun zach-like game is making it more like a puzzle game than a simulator, and as such the level design is paramount. Ideally someone should have a good feel for whether they're likely to pass the level before they hit run. If it's a tricky problem, they should be able to get a sense of how to debug / fix an issue from the feedback they get from the output.
You don't want your players to end up trying a bunch of random things to try to brute force their way through the problem, they ought to be able to reason about the solution.
You ideally don't even want to teach the maths, if even an expert needs to hand-crank calculations outside your environment then you're going wrong somewhere. Try to find a way to visualize the current or predicted states in a way that those calculations can either be abstracted or represented in a different way, so that someone can play without doing the maths themselves.
An alternative is to provide a mini sandbox area within the problem they can play around to get a feel for the maths so they can break the problem down that way. If you can find a way to represent the calculations within that somehow, then that'd be a bonus.
I'm also building a circuit simulator (stuck on that for a while, tbh), although not necessarily simulating electronic analog computers, but rather the computational principles that could also be applied to mechanical circuits, or just numerical computation.
As for abstraction, I really like the approach of Pure Data, you can group some nodes into more abstract ones.
Digital computing has the benefit of abstraction. You don’t need to know Ohms law to think about logic gates. If I commit to analog computing, then I absolutely have to find a way to teach the math. The operations are no longer Boolean, they’re a combination of a bunch of different electrical equations. Even if it’s simple math, it doesn’t have the same ethos a programming game typically has.
Does this even sound interesting? Would you play it?