Perhaps. I could also see ramped up lauch cadence allowing for design with standardized components or interfaces, as they know a service mission can go up every couple of years, so using a cheaper standardized component...
Yes, for sure.
Hyperbolic hypothetical: One battery maker could have a single modular battery that every satellite uses. That single battery might last a full mission for low-demand one off satellites. You might need a few of those batteries for a 4-5 year Starlink-like LEO, and maybe that battery set craps out after half life. A bunch of those batteries might only last 1-2 years on a higher orbit satellite [that's in a worse space environment]. Because of the production volume that battery would be mad cheap and readily available and predictable, and so one can imagine a scenario where its actually a better deal to keep replacing these lower-spec-than-necessary-but-crazy-cheap modules rather than every satellite using a [somewhat] bespoke full-mission-life battery design, as is done today. One could further imagine (as has more or less been suggested here already) that used modules with life/performance left in them could be reused on lower demand satellites. It sounds fantastic. Kittens and rainbows.
The main point is that bringing even individual parts of that hypothetical to bear requires clearing many fundamental roadblocks...many of those will take so long to clear that they could very well be OBE'd by other variables in The Big Equation (like decreasing launch cost).
A great thought experiment on that one is Tesla batteries. One could imagine a scenario where Tesla could repurposes degraded long-range batteries into standard range vehicles. Or, one could imagine a scenario where Tesla over-currents battery technology such that the user gets higher acceleration/charging without a larger/heavier pack, but the battery requires replacement every 10k miles. There's plenty of knobs to turn there, but it's worth trying to noodle-out the actual value proposition to Tesla for those kinds of scenarios. They're not easy to close, nor are they easy to see a path to ever close.
In the context of this thread, Space Stuff is similar.
IMO, the primary focus in the near and mid-terms (= for the next few decades) surrounding this topic at large is going to be safe disposal rather than equipment salvaging/re-use or serviceability/swapping. With Starlink, SX has step functioned both the problem and the v1.0 solution here: They're of course hucking up an unprecedented number of satellites, but they're also taking the unprecedented step of making them all fully demisable (part of that is genuine good-stewardship, and part of that is simply a function of meeting the statistical requirements for human casualty). v2.0 IMO is going to better address longer term risk of satellites turning into space rocks and Kessler-ing the whole party, as well as the environmental risk of atmospheric disposal. FWIW SS could actually be an enabler in this space where, practical roadblocks aside, it could bring lifed-out mass back down to earth for more controlled and efficient disposal/recycling.
