Today, there is a huge gap between high-end and low-end VNA.
And we should remember to be happy with that "low-end" if you compare to 10 years ago
Siglent and Rigol understand that, and full the 1500$ - 5000$ range (more than 3000$ if you don't hack), deserted by high-end company doing high-end stuff.
But there is still the 500$ - 1500$, and the question is, why ?
This range is useful to a lot of tiny company that design IoT or consumer-range product. Today, a lot of designer copy older design, and say "it should be okay".
But what if you want to check your RF path between 100MHz - 2400GHz ? You "need" a VNA.
And if you want to check what you're actually transmitting ? You need a SA.
This is why the 500$-1000$ range is so important : it's a no brainer for small company, even if it's one time of use week per year.
More than 1500$ ? Or 5000$ ? You're gonna argue, "Why now?", "Can we do without?", ... And everyone think you just want a new toy (even in the R&D team).
In my last job I had to wait for some BIG problems (certification failed) to make them buy a signal analyser, and even at this point it was after WEEKS of arguing.
Can't wait to test the LibreVNA, and see how it compare with the NanoVNA V3 next year