(A) the potential lifespan of these satellites are?
(B) Why has this not been done to date? I assume SpaceX expects that it'll be relatively affordable since they can reuse rockets.
(C) Will this impact the launch capabilities of future rockets (4500 satellites are a lot)? My assumption is that the low earth orbit will likely aid in destroying those satellites as opposed to leaving them up there at end-of-life
B) It's a triple play of radically reduced launch costs, rapidly dropping space hardware costs thanks to growth in the market, and improving RF tech to greatly increase the bandwidth such a constellation can provide. Ubiquitous connectivity is a great example of a technology that's going to happen because of the high level industry trends. The interesting question is whether it will be a SpaceX constellation, OneWeb constellation, or an atmospheric solution like Google's Loon or Facebook's Aquila.
C) 4500 satellites is more than we currently have in LEO total. However, LEO is low enough that atmospheric drag deorbits idle satellites relatively quickly. SpaceX's constellation is designed to deorbit on its own after their 7 year life.
Which is a very recent phenomenon, and still not fully materializing – SpaceX' Falcon 9 already cut costs in half in its class without reusing anything. Actually reusable Falcon 9s and the planned competitor products will drive down costs further still.
For starters, since they will be in a low orbit (to reduce latency), each satellite will have a fairly low utilization. They will usually be flying over territory with no or very few humans within range.
That would be, what, $9 billion? That sounds like a lot, until you realize they can spread that cost over 2-3 years and have millions (if not billions) of people paying for internet. This would replace cellphone plans too remember...
So that means SpaceX could probably charge a nice $100 - $300 a month and people would clamor for it. Seriously, this would probably make 10x return (easy) per satellite, if not more. Even 10 million people at $100 per month is $12 billion a year. Multiply that across a two to five year lifespan of a satellite and you'll have a very profitable business.
>millions (if not billions) of people paying for internet
You're vastly underestimating the challenge in attracting millions (let alone "billions", that's so ridiculous it's not even worth commenting on) of costumers.
>This would replace cellphone plans too remember...
You can't just hope to replace 4bil+ users in a snap of the fingers... Even if you had a hypothetical magical product that was cheap, available and revolutionary, flipping 4,500,000,000 clients would be a grueling task.
>could probably charge a nice $100 - $300 a month and people would clamor for it.
Why?? Who would be willing to pay 300$ a month for internet?? Maybe wealthy people in developed countries would pay that for the convenience of replacing home+mobile internet, but you wouldn't ever reach the amount of users you project with those kids of prices.
Right now people are paying Comcast $100/month for 25mbps, and Verizon $100/month for 5mbps at a cap of 10Gb...
I grant you you wouldn't hit a billion users at those prices, but you wouldn't need to at those prices. I guarantee internet like this will start at a premium, maybe $500/month. Then as more people adtop it'll quickly drop to something more inline with $50/month (or less).
You say I'm exaggerating getting millions of users. However, Id like to remind you Tesla preordered how many cars? Google Fiber had how many customers switch over? This is a game changer, and one that makes financial sense on both sides. Honestly, this is going to be like the Model T or iPhone, they'll easily have a million users their first year, ten million the next, and so on until they reach market saturation.
It's also a good bet Musk has done the math and thought about future math as well.
This won't just be profitable. It has the potential to be a political game changer.
After a while it's going to be possible to start running servers on satellites, to move them out of national jurisdictions.