Thank you, I did fast read some forum post about TDR and that person does not sound like he knew what it was about.-)
Mr. EEVblog has talked about the small "pocket" scopes being a toy, compared to a real scope, do I assume the same about the nanoVNA? Or is it really working fine?
NanoVNA allows to see TDR impulse response. Basically it's just inverse FFT transform of a complex reflection coefficient taken in a frequency domain (this is what the VNA actually measures). In such way you can transfer reflection coefficient measurements taken in frequency domain into time domain and can see how reflection coefficient was changed over time. If you want to see TDR step response you can use MATLAB script that I posted above.
The only con of NanoVNA is too small bandwidth, in time domain it leads to small resolution. It good enough for cable and long transmission lines, but too small to see all details of transmission line on PCB, because it's too short. This is why I recommend to use VNA with 3-6 GHz bandwidth, it allows you to see more details.
For example here is TDR impulse response of two coax cables connected in a series with open end. It is taken with NanoVNA. The first cable is 0.5 meters RG316 and second cable is 1 meter RG58. RG316 has SMA connectors. RG58 has BNC connectors. There is also SMA-BNC adapter between cables.
You can see a small impedance mismatch at about 4.8 ns, this is impedance mismatch on SMA-BNC adapter. Also you can see a high impedance mismatch at 15.08 ns, this is impedance mismatch at open end of RG58 cable.
As you can see NanoVNA is good enough to start. And if you will need better resolution then you can select more expensive VNA.