I'd like an explanation of what is being shown as the VNA curve does not look correct to me. I'm a retired PhD level reflection seismologist, so I've spent a lot of time with the Fourier transform both analytically and in discrete form computations. Over the course of my career I had to maintain over 2.5 million lines of other people's DSP code. It looks like a bug to me.
In particular the curve labeled VNA transform is not causal. However, it is neither minimum phase nor zero phase. Bode proved that for a physical system it must be minimum phase, in which case it should closely match the analytic result. As it is not symmetric, it's also not zero phase. Which raises the question of what is it and why?
How was the figure generated? What was the frequency domain input? How was the negative frequency part of the series set up when doing the transform?