These are good points, but I still think it's technically feasible to make non-aluminum satellites, and reasonably economical even if it costs somewhat more. If the only alternatives are giving up on large satellite networks or destroying the ozone layer, then switching away from aluminum seems like the way to go.
I'm not claiming steel is the best option though. Other metals, or even non-metal materials as mentioned in other comments might be better.
I don't think needing new components is really a downside except in the very short term. With much cheaper launch at much higher volume, we're going to be doing a lot of that anyway. We're going to have all sorts of new applications that weren't economically feasible before.
True, it's definitely feasible to make non-aluminum satellites. I do tend to go get up on my systems engineering soap box when people just talk about launch cost as it's a conversation I have too frequently in my job as well.
The mass consideration has certainly changed things in general. One of the things we have seen is a shift from aluminum honeycomb panels to just machined aluminum plates which weigh more for the same stiffness but are easier and faster to manufacture.
Steel might also drive more spacecraft to have to do a controlled re-entry (something the Starlink satellites are not currently designed for) as with the higher melting point and more mass it is more likely to survive re-entry. Although it does have a lower specific heat capacity than aluminum so I'm not quite sure on that one but usually the things I have seen that we expect to survive re-entry are titanium tanks and large glass mirrors. They also have the benefit of being in the center so there is a bit of an ablative shield.
I'm not claiming steel is the best option though. Other metals, or even non-metal materials as mentioned in other comments might be better.
I don't think needing new components is really a downside except in the very short term. With much cheaper launch at much higher volume, we're going to be doing a lot of that anyway. We're going to have all sorts of new applications that weren't economically feasible before.