In Townes (1983), it is proposed that the commonly accepted view that SETI should operate in the microwave might not be as robust as it seems.
Before this paper, most searches were proposed to be performed in the microwave region, but other regions of the EM spectrum can be shown to be more valuable when other considerations and conditions are used. For example, IR may be better than the microwave region if one considers the use of photon counting instead of linear amplification.
We do not know what design parameters ETI considers important for METI, so we should be very cautious about limiting what frequencies we search for. The microwave region is good in that it can be searched in “right now” (in 1980s time).
Shorter wavelengths could also be better if the geometric directivity of their telescopes can be utilized.
The paper is notable for looking at previous assumptions in SETI and trying to remove ones that may be unnecessary or ill-motivated. I also appreciate that it stresses that we should not get too confident in our guesses for what frequency ETI will use.
The paper still has practical considerations which will always be dated. It does not do a deep dive into the physical upper limits of transmission. This is not strictly a problem, but we now have technology that far surpasses that of this time, so the arguments for search recommendations are now outdated.