so what is the solution to the routing problem?

It's been years and has gotten progressively worse; it feels awful not being able to play due to the spikes in ms.
Once the instance is created with the routing issue, there is absolutely nothing that will save it, rendering the instance completely unplayable and resulting in having to abandon the map entirely.

What is the fix?

It's not shader caches(lmao), switching servers nor a vpn. The instance remains completely unplayable.

Pressure GGG (which is happening anyway)

bad nodes has always been a thing but it was relatively rare like 1/500 maps something like that you'd have to bin because you got shunted onto it.

The truth is their service provider is screwing them, but you can't just change provider at the drop of a hat so solutions can't be instantaneous.

Basically there isn't a solution aside from wait for them to fix it, which sucks because its really bad right now when we need it to be stable and capacity requirements are at their highest.