The problem you run into, if you're talking RF, is that inserting a switch
degrades Q by its series R, and pulls frequency by its shunt C. You can
design to this somewhat. But particularly with inductors switching the
coils is liable to make them worthless (you're talking ohms on the switch
vs tens of milliohms in an IC-scale inductor, less in a wire-wound).
If you have a limited quantity of frequencies of interest it may be
better to make the tanks the "element" and switch outside the core
LC loop.
If you don't care about compactness or solid state implementation
you could (as makers of high end load-pull test gear still do) use a
moving core inductor with stepper motor control, similarly you can
still find mechanical variable capacitors that could be fine-positioned.
Of course these are bulky, prone to microphonics and so on. But you
want to think about your "values" and what will be practical for you.