Is it really wise for Space-X to put 11,000 satellites into space for a new way to get internet

Latency and bandwidth. Latency could actually be better than fiber optics on the ground, since light is faster through a vacuum than through glass. A few hundred miles isn’t much in any case.

The improved bandwidth comes from the overall large number of satellites (each one only has to serve a smallish portion of the globe), and from their increased ability to use beamforming. I’m not sure what their beam spot size will be, but it could easily be less than a mile, as compared to the hundreds of miles for geostationary satellites.

Space may be a big place but Low Earth Orbit (LEO) aint so big and it is getting dangerously crowded. I present to youthe Kessler Syndrome.

And as this recent Wired bit goes into, keeping track of the increasing number of trackable objects is getting harder to do … and the many many more objects too small to track (even paint chips) can also do enough damage to start a cascade.

Yes, the concept is that these many objects should be designed to deorbit naturally … but the risk of a one collision leading to a catastrophic cascade is real, not tiny, and increases as LEO gets more filled up.

Musk thinks that his Kessler fleet will have no impact on astronomy.
He really needs to build some good will on this–I suggest that he goes out in public and shakes lots and lots of hands.