You have calibrated over too wide of a frequency range. I did measure the coax to the antenna using the VNA and a pulse generator and transform and velocity factor.

I did disconnect the coax at the Butternut antenna and shorted out the center conductor and the shield to eliminate static build up before applying the VNA to the Butternut's 75 ohm matching stub.

I really don't think that I did anything wrong setting up the VNA, so I'm betting on antenna problems. If I scan the entire ham band, I still get the 1.2 SWR!! Either this is the best antenna that I have seen in my life, or I am doing something wrong here. On all the bands I get an SWR about 1.2 across the whole band, no matter which band I choose. Since this antenna covers 8 bands, I set up the start and stop frequencies to cover each band, ie 14 - 14,5 MHz, start and stop, for the 20M band. I calibrated the VNA from 50KHz - 40MHz, O, S, L. If I set the format to RESISTANCE the nannoVNA returns 50ohms from 10kHz to 1GHz and above that frequency it returns a resistance from 49.8 to 50.2ohms and a reactance of few mohms along the full span range.I set up my NanoVNA to measure SWR on my Butternut antenna. Note that the SWR graph seems to don't have any trace, but it's because the trace is on the bottom axis. I also tried to fill with FF the whole flash memory and reloading this version but the result is as you can see below: Note that activating on other traces the return loss and the Smith chart those formats work great.

While before this firmware update I had it working good, now this format reports always 1:1.00 despite the DUT attached. The strange behaviour I had is about the SWR format. But that is just a question of preferences, mine could be different from yours and it's not an issue at all. I calibrated the nanoVNA using the average and it seems to improve the calibration indeed, even if I think that it should be better if the average flag was ignored during the scans. I downloaded version 0.4.4 but I encountered a strange behaviour.

Nanovna swr