Presumably to find out whether the difference lies with the filesystem or somewhere else?

If Linux is still much faster, even with the same filesystem, you have eliminated one variable.

Doing some profiling and system/kernel level analysis would be much saner, imo. What's the sense in measuring how some non-native filesystem behaves? In the end you'll be benchmarking how good is fuse-ntfs vs. in-kernel-ext4 and figuring it's slower... I say, profile some code and see how much time is spent in filesystem calls.

But then you'd be comparing NTFS implementations. The codebases are completely diffrent.