The system has been running 2 collectors, one that looks at W0MW.DYNIP.COM and one at W3LPL.NET. I have added another one to also collect spots from WA9PIE (hrd.wa9pie.net); it is running now. I see that some of the spots wa9pie provides are dupes with ones from the other nodes(this is to be expected); and some are different. Anyway, I will leave this running for a few days... please check dxdisplay and see if it contains more of what you are expecting now.
I have added a feature which allows you to select a DX Entity from a drop-down list, where you can pick the Heard From entity. This should improve the usefulness of the system to hams outside the US. Unfortunately I don't have grid selector information as this is not included in DX Cluster spots; I suppose one could infer that from the lat-long info, but I'm not sure how useful that would be due to accuracy concerns. The Heard-In feature works for the histogram and the history/trending screens.
Alex, This is one of the key but subtle issues in PSK reporting, wspr, and so forth. Many attribute the absence of reporting stations as reflecting "poor" propagation conditions to that geographical area while it's a "sample selection bias" (Google Heckman sample selection bias) because of the human factor you point out. It's an unbalanced dataset by design where there are uneven, even structural zero counts, in the basic design. Keeping that in mind, however, is what's key. That does not mean the data are not useful. Quite the contrary!
