This is an important point. Transit over StarLink looks like it will be faster than transit overland via fiber optic cable due to the speed of light being slower in fiber cable than free space.
That means that this isn't just a stunt, but an actual upgrade to our communications infrastructure.
Honestly, I’d love to see StarLink keep the latency advantage to themselves and acquire an HFT firm to extract the profits from the advantage. What are you going to do as a competitor? Build and lift your own comm sat network? The laws of physics are a hell of a moat.
If they went this route, they could run the compute at the edge; that is, in space, perhaps in another sat in a higher orbit. Thereby avoiding one trip through the atmosphere improving their latency advantage even further.
The real trick will be acquiring rooftop space at the exchanges for the uplinks.
Acquiring an HFT firm won't be worth it. Just sell it to the highest bidder. It's likely they will pay too much anyway (winner's curse) in which case Starlink just needs to sell it to the next highest bidder a year later. Much easier money (and less regulation) than trading yourself.
That means that this isn't just a stunt, but an actual upgrade to our communications infrastructure.