A well written article and I agree with your conclusions
However, you state:
Purely analog ISO increases do not cause more noise (easily enough tested empirically), because amplifier noise is minuscule compared to the other components of noise.
Analog ISO does cause an increase in noise, just as it causes an increase in the signal. An analog amplifier cannot differentiate between noise and signal, it amplifies both.
I suggest that a more accurate statement is "Purely analog ISO increases do not cause an increase in signal to noise ratio ... ". Assuming that elsewhere the result is scalled to the same lightness, there is no increase in the noise in the result.
Comment on Analog ISO
A well written article and I agree with your conclusions
However, you state:
Analog ISO does cause an increase in noise, just as it causes an increase in the signal. An analog amplifier cannot differentiate between noise and signal, it amplifies both.
I suggest that a more accurate statement is "Purely analog ISO increases do not cause an increase in signal to noise ratio ... ". Assuming that elsewhere the result is scalled to the same lightness, there is no increase in the noise in the result.