Wow, that was CLOSE!!!
Word of advice - ALWAYS test your cable cams from start to finish before taking your eyes off it and just expecting them to fly perfectly.
Rich beat me to it, but this is what I was typing when he posted:
I suspect the issue (not that it's really an issue) is the nature of the path that's generated through your points. One of Solo's best features is how it flies a smooth path between the points you set. To do so, it generates a curve that goes through your points, but will overshoot each one to smooth the path out. In the pic below, all 3 points are at the same height, but you can see that the curve overshoots each one. These curves will be more extreme the tighter the "corner" is between the points before and after a given point. Your points varied from close in to a ways out and Solo had to make somewhat drastic changes of direction to fly through them. Large changes in x, y axes (left/right/forward/back) can make the curve exaggerate smaller changes in the z axis (up/down). Also, the speed at which you fly the cables will effect how much it'll overshoot each point - if you're going faster, it'll overshoot each point by a greater degree.
Lastly, while the barometer is amazingly accurate at detecting small changes in pressure and so being able to maintain the drone at a fixed altitude, I think it's less accurate when trying to fly back to a precise altitude. Think of it like squeezing something between your fingers - you can maintain a pretty good, consistent pressure while you're holding something, but if you were to put it down, pick it back up and try to hit that exact same pressure, you'd probably be a ways off.
Which brings me back to my first point - always test fly your cables!