Hi,
On the original poster's use case, I'd suggest that GNU Radio is not an ideal or even rough way to understand filter response because it's not generally used to replace a swept spectrum analyzer.
A helpful person posted a video on how they characterized antennas using a broadband noise source from ebay (~$20) and a directional coupler (also ~$20). He used Spektrum, an RTL-SDR spectrum analyzer that can do sweeps. I copied his exact approach and it worked to (very) roughly identify unknown antennas. It's easy to modify the setup to sweep a filter to see its response.
On GNU Radio, I'm no expert but it's more for building real-time systems. It can do a few things an SA or VSA can do, but not all. It can playback files, but not scrub through them like an RTSA (e.g. Tek RSA, Signal Hound). It can be used to design and modulate carrier signals using an SDR, but it's not as convenient to pick up and use like an AWG or VSG.
What GNU Radio can do that test instruments cannot is this awesomeness: radar, space monitoring, cellular base stations, satellite ground stations... you name it. It's a wireless communications system-building platform - as a friend noted.
I hope The Signal Path, or W2AEW, or EEVBlog have a helpful review of of SDRs and GNU Radio and then help raise awareness about how they complement test equipment.
HTH