On the surface your question is valid: Why have 2 devices in series that "on the surface" appear to do the same thing (i.e., mitigate noise on the ethernet "signal"). Well, as it turns out (from my experience, and many many reports by others who have also tried) cascading multiple "high quality" switches in series has proven to provide an aggregate sonic uplift
that is better than either switch operating alone. Even though each switch (in this case, an eRegen, and a PhoenixNET) is completely competent on its own, the compounding affect of having them COMBINED (in series in a controlled environment) yields a better uplift than either one alone.
To see my notes of my experience, kindly refer to this posting if helpful:
Report Update (continued from entry #131 above): Greetings all. I have now done some initial testing of A-B-A Test 2, so I will re-iterate the digital baseline again for clarity: New 'Current State' of Digital Front-End: ISP Modem (JS-2 LPS) -> eRegen (JS-2 LPS) -> PhoenixNET -> Zenith Mk3 ->...
NOTE: IMHO, one key recommendation here (when trying this) is to be sure to give all your ancillaries (e.g., DC cables, PC's, LPS's, switches, etc.) adequate time to fully break-in and settle before making any final performance judgements. Sometimes these sensitive ancillaries take several hundred hours to fully break-in (e.g., their capacitors, dialectics, transformers, etc.). What may sound harsh at 25-hours, may smooth-out and bloom at 400-hours. That said, your first plug-in should be a good indicator of the direction you are heading before the "break-in" rollercoaster begins.
Hope that helps to answer your question.