I'm interested that Paddy saw a 16" lightbridge mirror dew up. Glass dews only when it drops below the ambient air temp. and any mirror 1.5" thick plus can not radiate out enough heat to do this normally, unlike the thin corrector plate on a SCT. I guess in theory a big mirror can dew up if it drops in temperature over the night and then the ambient air itself warms up with a change passing through, though this would be rare.
Keeping the fan blowing should hopefully prevent even this, as in the 16" Tridob.
Whereas a fan blows air in a fairly straight direction from the front, it sucks from a wide, diffuse area behind itself. If you don't believe me then try blowing a piece of paper from a distance. Now try sucking it from the same distance - see what I mean?
Sucking the air out the bottom of the tube would only work if the bottom was completely sealed. Even here, I suspect much air would pass down the tube past the mirror without touching it then out the back, though any air movement is better than none.
If you want as much cool air as possible making contact with the glass then blowing towards it is better.
Also, if you don't have the bottom fully sealed then at least use a baffle around the fan otherwise much air will simply recirculate from front to back.
Boundary layer fans only improve high-powered images temporarily while you are waiting for the glass to cool down, although having air blowing on the front as well as back surfaces of the glass would speed up the cooling process a bit.
For more info see:
http://www.cruxis.com/scope/mirrorcooling.htm
If you fiddle with the parameters, you soon discover the biggest jump in cooldown time happens when you add one fan of any reasonable power. After that you get diminishing returns.
This is why I only used one fan in the sealed mirror box of the Tridob and again only one (baffled) fan in the 22" Compact.
