This is a harder question to answer than it appears at first blush. As a historian of surveying, mapping, and spatiality, there are a few different levels on which an answer might operate. There's truly gross error (as in, landforms that either were absent on earlier maps, or thought to be present but weren't really there); errors in relative position of known objects or the appearance/disappearance ofsame over time; small errors of size and scale in persistent objects; and so forth. These must be matrixed against the kind of thematic information one is considering: topography (landforms, hydrography, etc etc), human geography, simple geolocation, or something else? The story of mapping is, in some ways, the story of getting the representation of the land right enough, and putting out information that's useful enough for whatever purpose it was collected.

In the most absolute sense of the concept, even with satellites in space, we still don't know exactly where everything is. The Earth isn't a perfect sphere, or even an evenly oblate spheroid; it's got all sorts of odd mass distribution and is slightly bigger around the mid-southern hemisphere than the mid-northern (as happens to us all around late middle age, perhaps). We have a very good idea of what is there and where it is, but satellite imagery isn't perfect, it's not always from the same angle or exact (known) distance from the Earth, and there's optical distortion to consider--just like there is with aerial photography.

If we back away from this highly precise standard, though, we find that satellite imaging did in fact highlight a lot of very serious problems with mapping. In my own subfield, most of the African continent was very poorly served in terms of precise mapping for many thematic categories. There was no ground-based triangulation framework for most of the continent, and there still isn't. The survey associates tasked with implementing title and commonage bounding systems in both Tanzania (F. N. Lugoe et al, see here-pages 7-10 give a summary of the uneven and problematic conditions, but there's a historical overview too) and South Africa (via the Chief Surveyor General's Office and the various provincial SGOs), to give two examples, discovered that entire districts hadn't been mapped beyond the basic landforms--the people, roads, watercourses, and so forth were all out of position, which created huge issues when they wanted to make improvements or consider land-titling schemes. Even though South Africa had good triangulation coverage otherwise via Trigsurvey/Surveys and Mapping/NGI, they put a lot of money into surveys over most of those regions (mostly former Bantustans) during the ten years after 2004. [edit: I'm not going to get into the politics of why; that's a whole different issue.] Tanzania still has the problem whereby positional fixes are too far away or simply unavailable to local survey teams. People who live there, of course, know where things are, but not their coordinates--just as it's not all that important for most of us exactly where our house "is" barring a property dispute, it's not for those locals either. But when a village or road is several kilometers out of position based on maps drawn up sixty years ago, and never corrected because nobody complained and it wasn't clearly wrong until overflight from space, that can create big problems down the line--because then you have to figure out where an object or structure actually is when you want to secure land rights, build irrigation networks, improve roads, you name it. That's where the issue of precision coordinates comes in. You can't fix it if you don't have the correct data at the necessary precision, but when all the data is wrong in different ways, sometimes it's easier to kick the can down the road than spend money you don't have to start all over.

The industrial world, and high-value areas outside it like mine concessions, big cities, or military theaters, are generally covered by cross-cutting webs of triangulation to control positional fixes and correct them if a measurement is erroneous. But in areas like rural Burkina Faso (another example), any triangles are huge in size and small errors in measurement can produce big discrepancies. The reason that remains so is because making up all the necessary trig points for reliably precise mapping takes a lot of time and money, and threatens to create disputes where vaguer understandings had allowed for flexibility of territorial claims. But at the time that satellite imaging became a useful tool, geographical knowledge in a great many areas remained dependent on imprecise traverses and flawed aerial photography, and so maps were noticeably incorrect then and a lot of them remain so today. That incorrectness didn't matter at a time when the state's only intervention was collecting per capita taxes, so it was never given priority before or after independence.

I'll see if I can dig up an online copy of the white paper on the South African case; it may still be confidential. I was there when they went looking for positional data [edit: on human geography], and we found that nothing existed except for trading station lots and municipal boundaries set circa 1903-1918--lines that had long since been erased by growth. As far as a description of triangulation, the belief in accuracy, and the limits of the "cartographic panacea," I still point people at the first chapter or two of Matthew Edney's Mapping an Empire on the great trigonometrical survey of India (Chicago, 1997).

Apologies if this isn't the sort of "incorrectness" you envisioned--I know there are other cases that fall into the gross-error category, but this is the sort that I work on.