I believe that most of us over-estimate their ERP by far (and I include also those guys with 'real' antennas).
I myself calculated from the antenna current and the height of my umbrella
antenna (18m) an estimated ERP of 200 mW. With my 1m
untuned loop I actually measured the ERP to be
Field strength measurements performed by Dick, PA0SE, Markus, DF6NM and myself also indicate, that probably only two or three stations within Europe actually reach the 1 Watt ERP line (considering all the difficulties of actually measuring ERP from a distance beyond near-field). G3WSC/p obviously is one of those stations and as I indicated earlier, their signal was exactly on my theoretical 1 Watt ERP line.
I am still surprised that with my actual ERP being that low, my signal still can be considered an average signal and allowed me to cover distances of up to 1300 km (one of the disadvantages of living in Central Europe is that there simply is no active LF station further away -- yet ...).
Even more I am surprised that with an antenna by far less effective
I was able to run my first Slow-CW QSOs over a distance of 402 km with
ON7YD last year with an ERP of (theoretcically, based on my actual measurments
and calculations) less than 0.1 mW (!!) ERP. This shows me, that, providing
we select the appropriate technique, the detection of estonishing weak
signals out of the noise is possible. The only other places I personally
know working with similar techniques are military (submarine communcation),
radio astronomers, NASA, and geophysical people (seismics and oilfield applications to transmit data wireless from the borehole bottom to surface) as well as our EME collegues!