I wonder if it is feasible to utilize filters instead of RF splitter/combiner and if anybody had this kind of experience in a past. As an example, assume that target application has intended frequency of operation as 1 GHz - 1.5 GHz and I can easily design LPF and/or HPF as 50 Ohm impedance on input and 100 Ohm on the output (or vice versa) with this tool : https://rf-tools.com/lc-filter/ In theory there should not be a problem to have these two filters to be joined as x2 100 Ohm to one 50 Ohm line as combiner and other way around as splitter - you got the point. Design does require filtering and splitters/combiners anyways, so why don't use this approach. Any takes on this topic and/or professional opinion?
Filter networks tend to be associated with phase changes when capacitors and inductors are involved. This may or may not be desired in your project.
Don't know if the same happens with splitters and combiners.
LC impedance matching circuits have typically lower bandwidth than dedicated splitter/combiner. In case of higher order LC circuits, phase matching would be questionable. Splitters like Wilkinson have also suppression of reflected signals that a simple impedance matching divider doesn't provide.
Thanks to everyone for valuable input! We will experiment with both - middle band single stage Wilkinson and will try 3-4 order Chebyshev calculated as 100/50 Ohm. It will be interesting to see how filters will perform - I probably will just cook up a small eval board.