The problem is this is also a myopic way of looking at things. What you should be looking at is also operational complexity. What's the current burden on your o

The problem is this is also a myopic way of looking at things. What you should be looking at is also operational complexity. What's the current burden on your org/team maintaining what you currently have? What about when you need to scale even higher? A lot of teams that think this way end up with really high oncall burdens and then never have the time to even iterate on their infrastructure.