Wasn't it lin8080 who wrote:

Hallo

Now, here I type my thoughts about.
Your water-hole brings the initial up.

Why not searching in a duple water-frequency? Is there ever a search or
a look at 2,8xx Ghz?

Or, wouldn't it be more intelligent to use math. oriented freqs? May be
3,14.., named PI, or e, or phys. constants. It is not known to me, ever
heard about such things. 

If you were to use such an approach, you'd want to cover multiples of
the H and OH frequencies themselves and multiples of (H+OH)/2. You'd
want to cover many multiples, like 1.5, 2, 2.5, e, 3, pi, 4, 5, 2*e,
2*pi, e*pi, etc. You'd also need to search some way either side of each
of these frequencies to allow for Doppler shifts caused by the proper
motions of the alien planet and of the Earth. You end up needing to
consider a fair proportion of the radio spectrum, so you may as well
scan all frequencies.

Several SETI projects scan the entire band where the background noise is
reasonably low. Below about 1 GHz there's too much noise from natural
terrestrial sources for the search to be feasible and above 10 GHz
there's too much noise from the galaxy.

It would be nice to run a SETI project from above the atmosphere, where
lower frequencies could be scanned, but the cost of one satellite would
be soak up the funds available for all the world's SETI projects for
several years. There's not a lot of money behind SETI.

-- 
Mike Williams
Gentleman of Leisure