I don't find it to be a workaround. The remote development stuff loads a bunch of projects and processes onto the remote host. I have some pretty small VMs that I run some older production servers in. Which is my principal use case for remote development.
Connecting to them this way brings them to their knees. That and seeing a host of new high memory processes doesn't give me warm fuzzies about the security risks on important remote systems.