The machine against the back wall opposite side of the door does not follow the perspective. You would not see the top of that box since it's above the horizon line.

I think you understand the idea. Your linework will look better and more confident with mileage; it's as simple as that.. Pretty soon you'll be amazed at how these types of drawings come together and the real fun starts!

Sounds right to me. Keep practicing the line confidence and mark quality exercises from earlier and slowly your vanishing point guidelines and boxes will look better and more confident, but it doesn't need to be a blocker before getting into drawing this stuff. Improvement takes time and practice.

That large shape describes the boundaries of the brow ridge on top, the edge of eye socket where it marries up to the zygomatic process on the lateral side, and the bulge of the eyeball and skin folds on the bottom, and then on the medial side where the eye socket meets the nose bridge bone (note: not the tear duct!). It's a construction technique used to wrangle the complex shapes of the eye into a simple, attractive shape first, in which the smaller secondary and tertiary forms can be drawn. It's a good example of working big to small to create accurate proportions and achieve the primary big effect first, then build upon that by adding smaller forms and details. Yes, the eyeball, lids and eyelashes are secondary, not primary forms when you're dealing with the entire head - features in general. Going in right away to draw the eyeball and lids without this construction will probably result in wonky proportions and much more difficult circumstances for applying shading and lighting the entire eye. I think with experience you can relegate that construction phase to your mind and go right for the eye while keeping the larger shape in mind.

1 - the entire top black part pulls off like a lid, so that the entire interior including the sharpening blades are exposed. You can just dump the graphite shaving out and then gently push the black lid back on.

I'd suggest trying to really drive home the box form and its plane changes in the torso (ribcage + pelvis). In your images, the large torso forms are flat-ish or represented with ovals. With mannequization, we want to really understand exactly how these large forms are oriented in 3d space, and the box forms Stan demos (building upong the robo-bean) will train that sense. I'd suggest following along with Stan's video demo, pausing frequently and copying his drawings. Then try it on your own. It'll sink in eventually.

Minor nitpick: Try to evenly light the photograph of your image and crop out distracting backgrounds, like the table. If you have an editing program, consider knocking grayscale drawings back to black and white so the tones are more neutral in hue. Your image has an orange-ish hue. Just a nitpick. Thanks for sharing!

Love how you deliberately show the planar facets in the image on the right. It shows you're thinking critically about directions of the planes on the form and which value to assign to each plane depending on where the light source is. This is the a good strategy I think.