You see it in solar energy forecasts as well. The last 2 decades have been an exercise in forecasters constantly underestimating solar energy generation. It’s actually quite embarrassing.
The problem I think is that the analysts in these areas are auto analysts, an industry that really hasn’t seen any major technology changes. The majority of cars sold today are internal combustion engines. Same as in the Model T.
And the analysts in the solar power field are energy analysts. Another field that hasn’t seen any major innovation that wasn’t another form of a fossil fuel for centuries (the exception being nuclear, but nuclear is kind of a bespoke technology for, if nothing else, all the non proliferation issues associated with it).
Is there a lore reason why none of the forecasts follow a logistic "S" curve, ie literally every other technology adoption curve?