Thanks for the suggestions everyone, I really appreciate it.
It does seem like there's a gap in the market here, but then again, I guess there just isn't the volume to drive competition and bring prices down. Clearly it's physically possible to make a suitable device with fairly simple hardware.
In the absence of an affordable direct replacement for the 4395A, I suspect we'd look at buying the Hioki. If it needs some additional software to run a full impedance sweep and plot the results, that sounds like a good project for a student on work experience.