Except it actually is. When a fully functioning satellite costs e.g. 200m$ and the launch is another 200m$ it makes sense to overbuild your satellite to 600m$ w

Except it actually is. When a fully functioning satellite costs e.g. 200m$ and the launch is another 200m$ it makes sense to overbuild your satellite to 600m$ with crazy redundancies etc to make sure it won't fail early on it's mission life. At the same time the big dumb booster concept like Ariane 5 and 6 makes tradeoffs like no upper stage engine restart capabilities which are ok for GTO missions but mean that multi-sat LEO or MEO constellations are impossible( you can't change the orbit and drifting is expensive).