So, is it safe to say that measuring checkpoint pages/sec is absolutely useless when investigating memory usage and pressure on my SQL Server.

High lazy writes/sec is the way to go - among other things.