pSemi PE4259 RF port behaviour when port is not selected

Let the below circuit:

The RF switch taps the signal between X4 and X5, and feeds to the RF2 (the normally connected) port of the PE4259.

When the Inhibit signal goes high, the RF switch connects the RFC line to the RF1 - which I have designed as grounded.

But - What happens to the RF2 port when Inhibit is high - and we have RF1 connected to RFC?

When testing the Panadapter boards, I found in the VNA testing Insertion Loss (S21 mode) between X4 and X5ports that when Inhibit is high, I have an non-neglectable signal loss in this line - I expected the RF switch, when the port is not selected, to disturb the signal to a minimum. See the VNA sweep from 10-100 MHz:

When the Inhibit signal is "Low", then I find the marginal Insertion Loss that I expected to have when testing S21 between X4 and X5. But, when Inhibit is high, there's this huge signal attenuation - which I expected to be undisturbed.

Reading the datasheet, it clearly states that PE4259 is an reflective type RF switch - Thus, again, I expected it to just reflect the signal back, without any termination - or grounding.

However, in a closer look in the PE4259 diagram in datasheet reveals the following:

Notice the red circles - For what I can see here, the default state for an unconnected RF port, is to have it grounded inside the chip - which would explain the signal loss.

Resolution

The always competent Gordon Brown from IntegralRF explained me (and I got this confirmed from Peregrine Semiconductor, the RF switch manufacturer, tips hat to Alex from pSemi tech support o/) that while the PE4259 is indeed an reflective RF switch, there are Reflective Open and Reflective Short subtypes - And the PE4529 belongs to the reflective short RF switch type - which causes the unselected RF port to be in short to the ground state.

In order to counteract it, I have flipped the RF Switch - and as a bonus point, taking advantage of the switch design, I was even able to get rid of a cap.

Now the input line (which should never be shorted) connects to the RFC port. During normal operation, RFC is now either normally connected to the RF2 port, forwarding signal to the amplifier - Unless when I get the Inhibit signal, which flips the RFC port to the RF1 port, now left floating - effectively, reflecting the signal and in insertion loss mode. See the below circuit: