The ellipse on the farther end of the cylinder, by virtue of being physically farther away from the viewer, is going to be seen as being wider than the end closer to it. It will also, by virtue of being farther away, be physically smaller on the page. By how much that far ellipse gets both smaller in scale and wider in its degree, is based on just how much foreshortening you're trying to convey.

As you make the degree shift and the scale shift more extreme, it shows more foreshortening being applied to the form, and therefore tells the viewer that there's more to the length of the form than what can strictly be measured on the page.

What we shouldn't see however are cases where the far end is dramatically smaller in scale than the closer end, but where it's still very similar in degree - because this would convey that the cylinder has both shallow (according to the degree shift) and dramatic (according to the scale shift) foreshortening at the same time, which would be self-contradictory.

As to your revisions, unfortunately you only appear to have extended the boxes' own edges, and have not applied any of the lines for the ellipses. Each ellipse should have a defined minor axis line, as well as two contact point lines, as explained in these notes. You applied these line extensions correctly when going through this section of the challenge the first time, so I do not really see why you would have neglected to include them here.

While your linework here is considerably more fastidious and far better planned/executed than before, which was the main issue I wanted you to address, I unfortunately cannot trade one mistake for another. Where previously the issue was that you got sloppy with your linework, and rushed through it, here not following the exercise's instructions in their entirety (despite knowing better) is simply another form of rushing.

As such, I will have to ask you to complete the 35 cylinders in boxes once more.