This is the result of about 6 hours of WSPR in the reliable 40m band, from 04:00 to 10:00 local time, running 10W, 10% transmit cycle and a Windom antenna. The north-west / south-east patterns seem to be predominant, but we must also take into account the fact that there are just a very few WSPR stations in Russia, Africa or South America, some of them were active (like PY2FLP) as I could see in the WSPRnet.org spot database, but no spots for me.
Then there is also the antenna's takeoff angle, wich makes the wave "land" at a certain distance from where it's transmitted; lower angle means your signal will go further, so the fact that I got alot of spots from the East coast of the USA but none from the West means my takeoff angle should be a bit lower if I want to work that region.
This can also be observed by looking at the closest spots, OK3SAM, IK3NLK or USP041 all being within 800-1000Km from my QTH. The way I picture it, most of my radiation bounces off the ionosphere pretty close due to a high takeoff angle, lands for the first time in the middle of Europe and then goes for a second ionosphere reflexion only to land in the middle of the Atlantic and partly on the american East coast. This tells me I need to raise my antenna higher if I want long distance contacts in the 40m band, wich also confirms my EZNEC simulations.