Besides niche scientific applications, this is mostly about extremely large constellations. The classic LEO, roughly 500km-1000km, is already quite full, and p

Besides niche scientific applications, this is mostly about extremely large constellations. The classic LEO, roughly 500km-1000km, is already quite full, and placing a lot of new satellites there increases the risk of triggering a chain reaction of collisions with orbital debris (Kessler syndrome). At these altitudes the atmosphere is thin enough that even small pieces of debris are long-lived, thus potentially creating a long-term problem for all kinds of human and non-human spaceflight. In the VLEO band, on the other hand, the thicker atmosphere makes maintaining orbit more difficult, but also makes debris very short-lived. Due to this, it may be reasonable to risk placing a million-satellite communications network somewhere between 200-300km, which would be utterly impossible further up. Data rates, latency, mesh size, etc. also improve of course.