I hate Bode plotters that work with that stepped frequency method. It not only takes forever to cover the frequency range, but by skipping between discreet frequency steps you can miss resonances/strange behaviors that might happen to be in between the steps. Real continuous swept sine or full band "white noise" frequency response is sooooo much better and there is nothing keeping an instrument like this from doing either method.
Can you determine phase response from a white noise signal source?
Edit: figured it out... http://www.dr-jordan-design.de/Download/Examples_filter_butterworth.pdf
The answer is yes. HP had at least a couple of instruments that could do this the HP3562A and the HP3563A. They are FFT based. I have the control loop of a switching power supply using a noise source. I had to do this because the power supply could not run for a long enough time for swept or stepped frequency measurements.
With these instruments you can use averaging, so after each measurement you would get better and better results until the result stopped changing.
For a frequency response analyzer, FRA, to be useful for switching power supply measurements, it must reject all frequencies except the frequency being injected. The scope might use FFT to do this.
If Daniel is reading this, I am very disappointed that the FRA (Bode plots) is available for the DSOX3000 T but not the A series. The product is DSOX3PWR