I have three G2s which I am testing with large antennae under my roof (Using multiple antennae and nodes in the same location ). I also have sometimes other nodes with smaller antennae in the same building which pick up nodes before the G2s (see link above).
I also tested the G2 with and 8 dBi antenna (in a magnificent location above the Vale of Evesham) against the T-echo with a Taoglas antena (Taoglas FW.86.B.SMA.M, ISM, 850 … 890 MHz, 240mm, 3.5 dBi, SMA male,
https://www.distrelec.de/antenne-ism-850-890-mhz-240mm-dbi-sma-male-verschraubung-taoglas-fw-86-sma/p/30285131 Distrelect Art. Nr. 302-85-131 Manuf. Teilnr. FW.86.B.SMA.M ) and found the T-echo remarkably similar in performance for node info detection as compared to the G2 + 8dBi antenna. ( Node info detection is much easier than message sending and receiving.)
My only guess is that a high dBi antenna can miss signals due to the narrower beam.
I found the antenna that came with the G2 to be absolutely sufficient as compared to the fancy ones I am testing (see link above).
To make a comparison one has to be very careful that exactly the same location is used. A few feet different and things are different due to position. Also the time you wait for the node list to replenish after a node database reset is large, about an hour or two. All my G2s receive node info at the same time in 99% of the cases even with different antennae.
I have not managed to stress test the message receipt and compare G2 + fancy antenna against a simpler node and antenna. But when on the road ( see examples
in here Car node with external antenna ) the G2 is sensational as compared to other nodes, at least it is always in the traceroute.
I would welcome any other ideas as to what protocol to use to compare node performance other than hearsay as I have given above.