Re: “. . .having the attachment point of belt to frame at the sides means that your hips cannot move up and down in their normal manner . . .”
The Stephenson design solves this problem by attaching the belt to the frame at only three points: At the small of the back, and at the points of the sidearms, furthest from the pack, forming a triangle. So each side of the belt is free to move independently up and down as one walks:

And each side of the belt rests on each hip bone (the ILIAC crests, not the hip joint), so the weight does not rub against the hips. The attachment at the small of the back is recessed, so only the belt, not the frame, contacts the back.
The attachment at the small of the back is a key element of the design, as it keeps the center of gravity over the lumbar area, and not further backward, causing the pack to rotate rearward, and place much more pressure on the shoulder straps, as was the case with the Jansport D model packs. and some others with sidearms.
A number of methods of attachment at the small of the back have been tried, including fittings that rotate 360 degrees: but only slight freedom to rotate at the small of the back is needed for the belt to rotate freely with the hips, which is actually more desirable to assist keeping the pack upright; and more elaborate attachment methods are not needed.
A modification allows the arms to rotate freely, and snug the belt as tightly as desired against the hips, as on this butterfly frame with a suspended mesh backband:


But using the Jansport fittings for the sidearm hinges requires 5/8″ tubing, too heavy for a really light frame; so in progress is a butterfly, or hourglass-shaped frame from prebent Easton 340 tube, which is less than half the weight of the Jansport tube. Roger Caffin was kind enough to use his homemade ‘Rolling Jenny’ device to prebend some Easton tube to the desired amount for the side arms and hourglass part of the frame.
Another lesson was the need to keep the sidearms several inches up from the pack bottom, again to limit top heavy rotation of the pack sideways. Using the common feature of bottom side pockets for water bottles also moves the center of side rotation upward, further helping to keep the pack upright..
Still another lesson in going to lighter tube was that the cross bars behind the belt must be strong and rigid enough to keep the sidearms from opening slightly so that the snugness of the belt is lost. A front buckle like Stephenson’s will solve this; but making the belt keep snug without a buckle makes for much greater comfort. .
The greatest thing about this design is that with a contoured hipbelt of ordinary width, slipping of the belt is a thing of the past, and all of the pack weight is transferred to the hips, with none on the back. The greatest drawback is that the frame must be more tailored to the wearer’s hips and body size, so while great for MYOG, probably not marketable. But there is more freedom for variation than initially thought.
Will post on MYOG when the frame is done, with more photos.