Both of these will compile to the same set of MSIL instructions. There may be some semantic difference at the IL level between the cleanest F# algorithm and the cleanest C# algorithm, but it wouldn't be significant.
As JaredPar said, both are equally capable of efficiently traversing large data graphs.
