> Satellites in higher orbits do a lot that can't be done in LEO
This isn't really true anymore. Yes there's a lot of legacy technology still in use and even still being launched, but there's nothing in MEO or GEO that can't be done in LEO with today's technology. Doing it in LEO requires more satellites and better radios, but you get better performance.
The risk of collisions may be lower but the consequences last thousands or millions of times longer...
The consequences of "there's some debris in a little-used MEO orbit for enough decades for it to take for somebody to be bothered to deal with it" are a little less drastic than "the world can't use satellites for 5 years" with the kinds of LEO crashes you'd be encouraging by replacing single satellites in higher orbits with dozens in lower. So even if the unit economics of replacing a geostationary satellite with a constellation large enough to maintain that continuous line of sight from a 500km orbit were acceptable, it would still be the exact opposite of a safety-enhancing move.
This isn't really true anymore. Yes there's a lot of legacy technology still in use and even still being launched, but there's nothing in MEO or GEO that can't be done in LEO with today's technology. Doing it in LEO requires more satellites and better radios, but you get better performance.
The risk of collisions may be lower but the consequences last thousands or millions of times longer...