I don't think anyone can provide a percentage because there are a number of factors, like how the memory is exercised, that will determine that.
That said, my personal experience is that you'll never see a difference using real-life workloads unless you're doing some VERY specific benchmarking. So I'd buy what's more economical. At the end of the day, all applications read/write to disk and 1ms difference in storage latency will mask any potential gain from the faster memory.
Technical Marketing, Performance Specialist, VCDX
@vmMarkA, virtualizationeh.ca