Starting with your first section - the cylinders around arbitrary minor axes - you're demonstrating good control over the ellipses in general, in terms of keeping them tight, confident and evenly shaped. The issue I did notice however was that while you do have some variation in the degree of your ellipses from one end to the other (some cylinders where the far end gets a little wider, some where it gets much wider), you appear to have no such changes in the overall scale of your ellipses.

Foreshortening - that is, the impact of perspective on an object, which is more apparent as a form gets longer - impacts both of these ways in which an ellipse will change from one end to the other. As foreshortening becomes more dramatic, and the viewer is told that this cylinder is in fact longer rather than shorter, the far end gets both wider in proportion and smaller in overall scale. This second point is what you're missing, instead ending up with some cylinders that are more inconsistent - with one aspect of foreshortening suggesting that the cylinder is longer, and the other suggesting that it's shorter. Keep both of these points in mind.

Now, I suspect that you just didn't really attempt to apply any variation in foreshortening, and instead varied the width of your far ellipses' degree because that had been touched upon in the notes. I generally leave it up to the students to pick up on these relationships between scale and degree shifts on their own. That said, with any such challenge where you're drawing many different instances of the same form, make sure you vary them in different ways - foreshortening included.

Continuing onto your cylinders in boxes, I feel you've largely done a pretty good job with this. The way this exercise is intended to work actually focuses much more on the boxes themselves rather than the cylinders. Where the line extensions from the box challenge allow the student to become more aware of whether or not their lines are converging consistently towards their shared vanishing points (ultimately allowing the student to adjust their approach and shrink those margins of error more and more over the course of the whole set), the cylinders here add a few more such lines with a similar goal.

The lines defined by the contact points and minor axes of the ellipses will align to the vanishing points of the boxes only when those ellipses actually represent circles in 3D space (assuming they're laid upon the various faces of the boxes). If the additional lines do align with the vanishing points, then we can indeed assume that those ellipses are circles in 3D space, and that the planes containing them are in fact squares (by definition). So what this part of the challenge actually trains in the student is a more intuitive sense of drawing boxes that have a pair of opposite planes which are proportionally square - a surprisingly useful skill to have as you get into lessons 6 and 7.

You've shown considerable growth over this set in regards to this particular skill, though it may have developed unbeknownst to you. The more you adjust your approach, the more you tweak that box from the get-go, the better your cylinders fit within it, and therefore the more square those planes ultimately end up being.

All in all I'm pleased with the improvement you've shown on this front, and while I feel more variation with the foreshortening in the first one is in order, you're overall doing a good job. I'll go ahead and mark this challenge as complete.