Not invalidating the rest of your chain of reasoning, since I agree with the idea this is likely to lead to eventual space based mid course interception….
While MEO satellites are able to do the job, they pose a greater “systemic risk” than LEO satellites do.
If a war goes hot with long range use of prompt global strike style hypersonic weapons against the USA or its closely allied military alliance partners like UK, Japan, Australia, Canada, etc… we’re talking about a scenario where the potential adversaries are going to be precipitously close to the use of anti satellite weapons, and potentially even nukes but let’s ignore them and focus on the satellites.
If you have a smaller number of larger MEO platforms (and they need to be larger to accommodate instruments that can to do their job from further away) your much more vulnerable to having large parts of your system knocked out. These larger platforms will also cost more, so you have less redundancy, higher costs, and are more vulnerable when your system will be needed most. The final systemic risk element is the post action debris risk… if a war goes hot and someone starts chucking ASAT weapons around, then we’re going to have a pile of debris, the sort of debris left behind by a kinetic kill ASAT in medium earth orbit is easily an order of magnitude more of an issue to other satellites for decades to centuries, it’s further away from all the existing space surveillance radar and optical systems on the ground making it harder to track the kind of small high velocity shrapnel that will have spread furthest away from the original orbit, which complicates the replacement of any destroyed space based assets as it will take longer to work out new safe orbits. LEO assets on the other hand have much shorter debris risk lifetimes, decades is usually the case as opposed to centuries, and LEO is much easier to track higher risk shrapnel debris down below 10cm, leading to better operational safety before and after a potentially satellite destroying hot war, meaning they are more likely to have a working system when needed and be able to replace it faster when damaged.
It’s a logical and reasonable argument on several fronts. However I do still agree with the assessment that eventually if the military has a sophistication LEO fleet of hundreds or thousands of Starlink style satellites, they are basically half way to the Star Wars “Brilliant Pebbles” architecture and it won’t be long until someone tries to build the other half of the architecture assuming they aren’t already planning this.