url : http://radiohamtech.com/page27.html
This post is an addition to the "pdf" article of the same name, available from the above url link. Below is a web search result for a beam antenna.
About 8,150,000 results (0.54 seconds)
Yagi gain vs number of elements
What I am mentioning here may not seem so, ao I suggest if your own test show the below is not so, then go by your own test results.
The above url lists a web page, the below url leads to the article, "antenna suspects"
url: http://radiohamtech.com/Antenna%20suspects%20v1%202%20publishing%20ok.pdf
From this article, a 50 ohm dipole has a 9dB loss. Most listed beam antenna are measured against a dipole as a reference measurement standard. Now there is zero or nothing wrong with this base line measurement, but I am under the thought that the beam antenna manufacturers are unaware that the 6 element 10·5dB gain above a dipole is perhaps more a 1·5dB gain overall.
The solution perhaps over this item, is to use a 50 ohm cross dipole. Page 7 of the pdf article, mentions a 50 ohm cross dipole with a near 3dB gain over a full wave wire antenna.
However there an upside.
The 6 element 10·5dB gain is a 6 element 10·5dB gain over the near 9dB dipole loss. If the dipole was a 0dB loss or gain, then the 6 element beam would a 6 element 10·5dB gain over the transmitter power output from the radio, not the unfortunate 1·5dB gain overall due to the unknown dipole losses.
Therefore, the suggestion and only a suggestion, replace the standard 50 ohm dipole with the 50 ohm cross dipole with its near 3dB gain over a full wave antenna.
The 6 element beam would then be, if correct here, a 6 element 10·5dB with a 3dB cross dipole boost, equating then to a 6 element 13·5dB gain over the transmitter or receiver performance.
At rough guess a 13·5dB gain is some time 20 times the radio performance. A 10 Watt transmitter with a 200 Watt ERP, and a receiver 20 times more sensitive.
However, do please go by your own measurement tests, to be sure for yourselves.