Drawabox purposely tries not to employ approaches that require us to plot everything back to specific, declared vanishing points - as these will often be way off the page, giving us a rather unpleasant dependency - and instead focuses on what we can work with given the space immediately around the thing we're drawing. So, we look at the lines present in the scene, and pay attention to how their convergences imply the vanishing points (rather than having them plotted out directly). This is the trend we've followed right from the beginning - we introduce vanishing points with the plotted perspective exercise, then shift towards estimating convergences by eye with a concrete vanishing point (rough perspective), and finally set aside concrete vanishing points when we hit the rotated/organic perspective exercises and the box challenge. This is a trend we hold to throughout the entirety of the course.

This does sacrifice a certain degree of perfection, but remember the core focus of what we're doing here. We're developing students' spatial reasoning skills, with each drawing being an exercise that forces the student's brain to work in 3D space, to manipulate that 3D space, and to understand how the things they draw actually exist in three dimensions, rather than focusing on the side of things that occurs in the two dimensions of the page.

This lesson, as are all the lessons, extensions of the simple matter of manipulating forms and then working within that framework of space. Falling back to stricter plotted perspective on the other hand is great for drawing everything with much greater accuracy/correctness, but it's much more difficult to both solve design problems and worry about that at the same time. So, learning to work without worrying so much about keeping everything so perfect and nailing your proportions is what will give you the breathing room to actually produce your own drawings and designs.

Keep in mind - and I believe I mentioned this when critiquing your cylinder challenge, as well as in the challenge video/notes themselves - the cylinders-in-boxes are an exercise that helps us hone our ability to estimate our proportions, and to help us develop a greater intuition for how a square in 3D space can be drawn regardless of their specific orientation in space. Being a prerequisite prior to moving onto Lesson 6 is intentional, as it gives students a stronger basis from which to construct more believable structures.

Lastly, the specific technique demonstrated in the "Constructing to Scale" video from Lesson 7 is presented as a starting point because it is at this starting point, prior to adding anything else that may predefine a specific focal length for the viewer of the scene, or other vanishing points/horizon lines, that we can make certain assertions that define these things for the scene. Meaning, when we lay out our first plane, and imply our first two vanishing points, we can assert that this plane is square by virtue of it being the only thing in the scene. The ellipse technique then allows us to adhere to that assertion as we transfer the unit measurement to other dimensions, creating a unit grid for the rest of the construction to follow.

This means that, as long as it's what we start with, we'll always start with something that is a square in 3D space. From there, it's a matter of maintaining consistent convergences with all of the other edges that imply relevant vanishing points, and working through the construction in a patient, step-by-step matter (without skipping important steps to save time). We'll inevitably decrease "perfection" or "correctness" as we progress, by virtue of being human, but the technique itself does provide us with a reliable starting point that holds to the core values of the course.