On a long steady run, the small difference in weight makes little difference (which is why tt wheels are optimized for aero and not so much for weight).
There is a shortage of testing on 'real' road surfaces and the last discussion on here about it turned a bit acrimonious.
Trouble is, what constitutes a 'real' road surface varies so much.
If you are happy on 23mm tyres, then fine, use them. Someone with less time in the saddle will find the ride very uncomfortable. The ability of the rider to 'float' on their bike when needed varies as well. Some people can maintain a good cadence and take weight off the bars and saddle, so that they are floating over the bike and not sitting there like a sack of cement being hammered by the vibration.
Edit: cross posted. Nice link, inc, and interesting to note that in Conti's tests on the 'perfect surface' drum, the fattest tyre has the lowest rolling resistance (and lowest puncture resistance).
Can't remember which bike mag it was, but i believe last year one mag had some riders give their impressions of the 'speed' of a variety of tyres, then compared it to the actual figures. The riders tended to rate 'harsh' tyres at high pressures as 'fast', even when their riding speeds were reduced and measured tests had shown that the tyres weren't the fastest.
