@MaxArt2501 @bramus The goal of the API is to be able to lay content on each side of the fold area (whether that's an opaque or transparent area) or avoiding laying out content on the fold area when it requires interactions (e.g. buttons). Basically, what you said making apps aware where the fold is so they can optimize or create new layouts.
Discussing with a few more folks I got convinced about the 1. I'll update the spec.
@MaxArt2501 @bramus I'm not sure I understand the "can't be reduced anymore". When the device is folded segment 1 + fold area + segment 2 = viewport size. If the window is a floating window, then neither on Android, nor Windows we support trying to intersect the browser window with the fold area to give you some partial segments. Segments are subdivisions of the total viewport.
The spec currently has `null` when the device isn't folded or it's a regular device.