I was typing my own comment only to find out I’m just repeating yours point by point. But I’ll add some more points.

The reason geostationary satellite transceivers were bulky is because of inverse square law. If you double the distance the signals goes down to 1/4. And since geostationary satellites were many times farther, beefier ground equipment is required. The upside is that it stays on the same spot in the sky so you only have to aim once.

Starlink, being a constellation of satellites, can be anywhere in the sky at a time, so its transceiver is a phased array, meaning it actively aims its signals internally to whichever satellite is passing overhead.

And why distance matters? First is bandwidth. Farther geosat means more signal degradation, meaning bandwidth is limited. Second is latency. It’s fine for passive data downlink like watching sat tv, but disastrous for time sensitive internet applications e.g. gaming.

One notable problem with Starlink right now is how all customers in the general area had to share bandwidth. So the more Starlink units in a general area, the worse the throughput. It’s like cellular network. The fix to this is more satellites to distribute the load, and Starlink is launching new batches every month or so, but you can only have so much. Also, competition is entering the market…soon…