Even with this benefit of the doubt, even stretching our suspense of disbelief as far as it can go, we are basically talking about a niche military application that inherently has a limited market value.
In other words, we can just assume this concept is possible and makes sense some level of financial sense for military, government, and high sensitivity use cases.
Well, find me a military contractor worth $1.5 trillion. Lockheed Martin is worth 1/10th of that.
Put Lockheed Martin and AT&T together and you’ve got about 1/5 of a SpaceX IPO target valuation.
Even if we make this assumption that the technology is marketable and has merit, it’s not like every company or government agency is going to want to switch to this technology. There are already a number of alternatives that can mitigate many to all of the risks that it solves.
The cost and complexity right now to deploy global services and their disaster recovery replicas to multiple distant data centers in the terrestrial world is already extremely low. Often, these features are offered as an off-the-shelf service.
California could sink into the ocean and my users wouldn’t even see a blip of downtime. Unless they live in California.
Heck, build some data centers underground in deep fortified bunkers if you want. That would be cheaper than launching them into space. They might even be easier to defend than satellites and space DCs because a foreign adversary can’t just launch a missile up in the sky to get to it.
Going back to not really suspending disbelief as much, I also think performance and latency is going to be an insurmountable problem. Right now as we speak Starlink residential service is about 10x slower than my home fiber connection at the same price. How is this technology going to compete with data center level infrastructure even in an optimistic scenario?
I can toss pennies per hour at Amazon for a relatively small VM with like 4GB of RAM and they’ll give that thing a 15 gigabit connection.