Thanks for the feedback.
I think even if I don't manage the image conversion to DDS, someone else will.
The good thing is that open free global sat imagery is available. Whether FG user experience with it could be pleasant enough is currently beyond me.
You may notice that the above example of Vienna Airport has one more problem besides the low resolution: it seems over-exposed.
The white areas representing some of the airport buildings and apron are actually not that bright. Real surfaces are gradients of gray, some not even so light gray.
Same for the north bank of the Danube river - it's brighter than it should be.
Why does this happen?
Turns out, the satellite cameras are not just taking RGB (red-green-blue) photos. Instead they pick up several bands - RGB specifically collected as Sentinel B2, B3 and B4, among other bands.
In order to get an image that we perceive as a photo, the 3 bands are merged. Here's for example one of the bands (blue):
The other bands (red and green) look similar. So the sensors seem to capture light gray surfaces as very bright and after merging we end up with these white areas.
Funny side effect is visible on the nearby highway: vehicles appear as elongated rainbow blobs. See how the merging algorithm created separate red, green and blue pixels. You can even tell the direction the vehicles are moving by the color shift: the front is blue, the middle is green and the tail is red:
It would be interesting if this over-saturation can be corrected or at least decreased a little. But that's a mighty project. Because here's the next caveat:
The cloudless global imagery we enjoy from some providers is created by selecting and combining multiple images of the same place. There are algorithms able to detect clouds and create a composite cloudless image out of several partially cloudy ones. The over-saturated outcome results of compositing of several over-saturated source images.
This whole story of how sat imagery is created is a little off topic but I thought it was interesting to share.