> very low earth orbits (VLEO) ranging from 340 to 360 kilometers
> Starlink currently has nearly 6000 satellites orbiting at around 550 kilometers
> its service ... with average latencies ... over 30 milliseconds at best, and double that at peak times.
What is unsaid in the article is how much of that latency is due to time of flight and how much is due to buffers and processing inside a satellite.
The stated fact that the latency "doubles at peak times" implies that some portion of the stated 30ms is not time of flight So how much is actually due to time of flight? One of the linked articles claims:
> 8 ms is the absolute minimum latency for a satellite at 550 km
When a satellite is exactly above head the round trip time at the speed of light is a bit shy of 4ms. This "8ms" must be some average over a minimum solid angle that still allows for an acceptable level of signal reception. It is a reasonable number.
Lowering to VLEO gives the above head latency of a bit over 2ms. Being lower, the solid angle for reception is less. They may actually need in total more satellites than at higher orbits. In any case, the average would probably be something less than the 2x from the 550km case. But, let's say "4ms" compared to "8ms".
So, by going to VLEO, the best they can do is knock off about 4ms from their "30ms at best". It seems that polluting VLEO will not even achieve "The biggest single goal for Starlink from a technical standpoint is to get the mean latency below 20 milliseconds," said Elon Musk.
There is some other reason for pushing into VLEO than reducing latency.