Overall you do appear to have done a pretty decent job of fastidiously checking for the correct alignment of your ellipses and the actual minor axes, and through the vast majority of these you have been correct. One thing that you missed, however, is the point I mentioned in bold in the homework assignment section. You appear not to have varied the foreshortening of your cylinders at all, opting instead to keep your side edges completely parallel (or as parallel as possible).

That is objectively incorrect - the foreshortening of the form (which is suggested by both the difference in overall scale of the ellipses on either end, as well as by the shift in their degree from one to the other) tells the viewer in relative terms how long the cylinder is. If the two shifts are more dramatic, that means the cylinder is longer. If the shifts are shallower, then they're shorter. If the shift is non-existent, then it suggests the cylinder has no length at all, which becomes contradictory.

Now, you did shift the degree of your ellipses, making the farther one wider than the closer one, but this must be done in the same measure with the ellipses' overall scale. Realistically there would not be a situation where one shift is shallow and the other is more dramatic - and therefore if the far end ellipse is much smaller than the closer end, then it should also be much wider, or vice-versa.

Aside from that, you have done a pretty decent job with this part of the challenge, and I'm pleased with the overall improvement of how you executed your ellipses. They tend to be shakier and less even towards the beginning, and that diminishes more and more throughout the set.

Moving onto your cylinders in boxes, I think your work here is largely quite well done. By extending the lines associated with the ellipses, we can check whether or not they converge towards the box's vanishing points. If they do, it means the ellipse in question is a circle in 3D space, resting within one of the box's faces - and therefore that face is proportionally square. As we continually repeat this process of attempting, then checking how far off we were, and ultimately adjusting our approach in the next attempt, we're training our instincts to better judge what proportions will register as square in 3D space.

I'm definitely seeing much more consistency in this as you work through the set, which suggests to me that your instincts are developing nicely. This will help you quite a bit in lesson 6.

So, I'll go ahead and mark this challenge as complete. Keep up the good work, but do be more careful about reading the instructions. When a particular point that was stated in bold is missed, it definitely suggests more care can be taken.