Are you comparing maps that have been post-processed and have been aligned using ground control or some other georeferencing technique? I would hope that they would align and look the same!
The question was "How good is that data" when it
hasn't been corrected and that's what I was referring to. I think we're all familiar with handheld GPS units and we take their accuracy (especially the vertical) with a grain of salt. These are the same receivers that are in the UAVs, unless you have a really expensive one, so you're looking at meters of error in the horizontal and dozens of meters in the vertical.
Here's a table that shows the accuracies when you have a complete GCP set and when you take them all away. When you have 0 GCPs, this project showed
2-4 feet of error in the horizontal and up to
62' of error in the vertical. Those values can swing wildly, depending on the satellite configuration and other factors. I've seen "pop-ups" fly overlapping missions hours or days apart and the satellite configuration changed enough to cause a 90-foot vertical change in where the drone thought it was. When they tried to compile the data, they didn't give the software enough "wiggle room" to correct for that and the point cloud had huge steps where the missions crossed. The point is that you can't trust the on-board GPS for anything other than relative values, at that time, and the way that data is collected is very important to the resulting maps.
Take the time to get yourself familiar with how your data collection can affect the final product and how accuracies are generated:
How Accurate is My Map?
How accurate is your drone survey? Everything you need to know.