If the interference is the result of a third order IMD product,
each 3 dB of rejection will yield a 9 dB improvement in the
third order intercept point.

A modest filter might yield surprising results, it isn't a linear
relationship.

Pete k1zjh

"Surprising results" in a lab maybe. Still far short of real. An
attenuator would tell you how much you really need. I used to use an HT
with a dummy load instead of an antenna from Mt. Wilson to be able to talk
into a box on Santiago Pk. Otherwise the HT couldn't even hear 500 WERP
on-channel from the tower I could see with my own eyes. Be aware that you
might only have 30 to 60 db of bolt on attenuation before case or cable
leakage takes over.

I used an Alinco 2m HT with a two section helical resonator outboard (most
portable solution). There was 3 db of insertion loss and 20 db of rejection
outside of a 3 Mhz window. This was a packet radio and resulted in a 10db
improvement in performance on-channel, but this was a home station on a 6db
stick 20 ft in the air. Would have been far short on a mountain though. I
aslo used that combination for T-Hunting in addition to a fixed 60db pad and
a switched 20/20/10db pad with double shielded coax and 4 el. Quad. My best
solutions was to find places to listen from that were shielded from the
major mountain tops.