How much electricity can plausibly be saved this way? And how does that compare to the energy required to reprogram and possibly renew all these signs?
About third of what was used before.
> And how does that compare to the energy required to reprogram and possibly renew all these signs?
That's cost of doing business and a one-time thing.
How much is that as a fraction of the total electricity consumption?
No idea how much energy it takes to reprogram/renew the signs. If they just turn them off, very little, I suppose.
A coder peddling the entire time required to adjust the timing would be lucky to power a single billboard for a night.
So I guess that's worth it, even if some additional hardware is required.
That's just energy they would have otherwise spent on making us buy more stuff.