One thing to keep in mind is economies of scale:

Cirrus: "new aircraft deliveries for the piston SR Series reached 380 in 2018"

By way of comparison, Ferrari delivered 9,200 cars in the same year.

Part of the problem with aviation has been that aircraft are boutique items and priced accordingly.

I'd imagine even a Toyota Camry would cost $150k if it was a one-off design.

> “innovation has a horrible ROI”

This may be so, but I don't think you've really substantiated it; you're just talking about part of the "I", without any reference to the dollars of "R". Sure, $1m puts a plane out of almost everybody's recreation budget. But for a piece of transportation infrastructure with high utilization, it might not be prohibitive. (I mean, a jumbo jet costs hundreds of millions to build, involves significant R&D budgets on each new model, and is still positive ROI, just.)

You said "you fly a bit faster, a bit safer" in a dismissive way, but how much is this actually worth? I could easily believe that making flights slightly safer is worth an extra $1m of capex; I think your original claim would be much stronger if you provided some more concrete numbers/analysis here.

> And that innovation is “just” glass cockpit, and some more advanced avionics. Not changing the principles - you fly a bit faster, a bit safer but still need a license and do all the preparation and checks. But it’s a bit more fancy. From 200k to almost 1 MM.

Nah. Avionics, etc, and what goes into a plane is cheap compared to the purchase price. Even when you talk about G1000, G3000, etc, they are not the primary cost drivers.

What's killed the price of general aviation aircraft is that very few units are purchased. In turn, prices go up; in turn more old planes are kept flying and fewer units are purchased.

Four seat aircraft without glass cockpits and without notable innovation are a few hundred K.

There's a lot of innovation now happening in the experimental/amateur-built category. Super-capable avionics, GPS everywhere, electronic ignitions and advanced engine intelligence/management. And for reasonable cost (relative to a new Cirrus and even a new Cessna). If you're willing to do the building yourself or trust another builder, you can get a very capable, modern airplane for under $100K that puts the certificated 1970s Cessna to shame in every measurable dimension.

You only pay the R&D costs once. All you're saying is that you have to put a lot into R&D, but that's all on the I side of ROI. It was the same with SpaceX, the I part was huge. The question is whether you can make the R part work. And if you have scale, then the R part moves differently.

It's the same with the first car you build. The first Tesla Roadster cost tens of millions to make. The second Tesla Roadster cost half that.

A new Cessna 172 SP with a Garmin G1000 glass cockpit and 180HP engine is 230,000$ that’s a significant upgrade over the 1970’s model for similar costs. Or you can spend 1/2 that on a The Cessna 162 Skycatcher with a 100HP engine and Garmin G300 Avionics.

100+k is expensive, but even a 162 is likely to be flying for 40+ years.