One of my favorite pastimes over the past few years has been playing a WSPR "game" I have taken to calling "radio limbo", i.e how low in RF output power can I go and still be spotted.  Since about 1540 UTC today I have had the WSPR RF output set at -10 dBm (0.1 milliwatts), hopping amongst the 40/30/20/17m bands.  At this tiny RF output level, the WSPR receiving "superstations" quickly emerged - ones like KA7OEI-1, N6GN/K, K9AN, KB9AMG, etc - still spotting my signal now and then.  No doubt many of them make use of the wsprdaemon software, and perhaps their owners are even active on this forum -  in which case I pause in admiration your efforts in achieving outstanding WSPR spot performance.  With my current setup in Michigan (grid locator EN73) I am not being spotted by the WSPR "superstations" in the northeast US simply because of terrain - there is a large hill (and the side of the house) in that direction that reduces the lower angle radiation....whereas the terrain to the NW through S is wide open.  Literally - as in the open water of a small lake.  Another interesting phenomenon at these tiny power levels is that the spots tend to disappear late at night - I surmise that the atmospheric noise levels increase to a greater extent than any enhanced propagation on bands like 40m.  All fascinating to me.

One downside of my "game" is that you can't advertise power levels <0 dBm in the WSPR protocol, so I have to keep track manually of what I am doing, rather than relying solely on database queries.  Tomorrow I'll begin reducing the power until there is but one WSPR spotting station still able to spot my signals.  Likely to be at an RF power in the 25-50 microwatt range, based on my past experiments.