I was doing some reading on noise improvement and stumbled on an article by Scott Hunt at Analog Devices. Here is the reference link: https://www.analog.com/en/technical-articles/11-myths-about-analog-noise-analysis.html#author
According to the first subsection, due to the Johnson noise, reducing resistance helps improve SNR. However, for applications such as measuring resistance, as V=IR, the gain in voltage supersedes the increase in Johnson noise. This leads to an improvement of exactly 3dB for doubling the resistance.
Does anyone know how he obtained the value 3dB, if the resistance value itself is unknown?