But solar has been getting cheaper and more efficient by leaps and bounds.

What would have been a poor investment 10 years ago, or even 5, might well be net-positive today, potentially even in suboptimal weather conditions.

I don’t believe the primary cost is so much the physical panel but the cost to engineer and design it into a roof, also the additional systems needing to hook it into the wiring harness. It’s a fun toy for some but has no real benefit for the many.

But even stipulating this, with the way solar is improving, it may very well not be the case for much longer.

If you can guarantee that, in moderately sunny weather, the solar panel on your car's roof can provide enough power to keep the car going at, say, 30mph indefinitely, that's no longer just a fun toy.

Now, that level of utility may still be a long way off—or may even never be possible!—but I'm not willing to write it off for good, given solar's curve.

ETA: sorry, realized I should unpack a bit why I think this is worth mentioning: Your GP post was expressing confusion over why people would study this; I think it's very valuable to continue studying it as solar continues to improve, so that we can understand just where we are in relation to that utility curve.