Personally, I think David has "nailed it".
Mauro asked could you post some pictures or sketches of this arrangement?
Here is a rough schematic diagram of how I have done it.
The main sheet is split into two parts, an upper sheet and a lower sheet.
The sheets run down to a block for each, at the sheeting position on the stern, and the two tails come forward from there, directly to the hand (or a double quick-release cam cleat mounted on the tiller).
The upper sheetlet arrangement for the two top battens is pretty obvious, and is probably the only way it could be done.
The lower sheet which controls the lower three battens is open to various configurations of sheetlet. The one shown here is simply what I hit on, and it worked first time so I never changed it.
I should explain, the reason for this was that I was confused by all the sheeting systems on offer, had no idea what to do, so as a temporary expedient I split the sheeting system into an upper and a lower, just to get started. It worked perfectly on this little boat, so I never "improved" it. As I have never experimented with different sheeting systems I have no idea if there is a better way of sheeting the lower three battens.
With this set-up, the two sheets may be trimmed as one, almost all of the time. The low-angle yard never sags away.
Mauro asks: The SJR has the same camber from bottom to the top and regardless to the type of sheeting system this problem did not materialize, am I right here?
I am not sure if I fully understand the question. I imagine, especially with the low yard angle, that a conventional sheeting system with no provision for anti-twist, would result in considerable sagging in the upper panels, and possibly a problem. I don't know for sure, because I never tried a single sheeting system, and for that reason never had the problem. I don't think being SJR has anything to do with it, and I will deal with the camber question further on.
.............
There are a couple of matters relating to this particular sail which might explain why this configuration works for this sail, but might not necessarily work so well for another. (I have noticed that there are almost as many different sheeting systems as their are junks, everyone seems to have their own preference).
Firstly, this is a five-batten sail. Most sails have at least six (regarding the boom as a batten). It would be interesting if someone who knows, could provide us with a proven sheeting diagram for a split sheet system for the more common six batten sheeting (including the boom).
The second matter is that people who know much more than me, often prefer not to take the sheet tail from the lazy block directly forward to the hand or the quick-release cleat, as shown here, but instead take the sheet tail back up to the boom, forward along the boom to a point near the mast, then down to the deck, and from there running back to the hand, or the quick-release cleat. This has the disadvantage of requiring more rope and more blocks (expensive) - notably much more friction - and double all of that in the case of a split sheet system. This might make a split sheet system not so desirable for someone preferring the more complicated sheet tail routing arrangement, which does seem to be quite common.
Mauro asks: would adding camber to the upper panels of a Johanna sail, change their angle of attach [attack] favourably? In other words, could a sail with same camber (8% in our case) from the bottom to the top, avoid the sheeting issue? In racing or costal cruising, optimizing the drive of the top panels could be an interesting feature...
I don't know, and obviously Arne is the one to answer that question, which I am sure he will do, when he is back on his feet and able to get back to his beloved computer. (Hopefully soon).
For the record, the SJR sail shown here has the same camber for all the lower panels, and a somewhat reduced camber for the single top panel. How this might have any influence on the choice of a sheeting system I have no idea. Also, I don't know much about special tuning for optimal performance etc. I have a feeling that my little boat sails best with a straight, vertical leech and no sag or differing angle of attack in the top panels. I don't really know for sure, but it just looks better to me that way, and it never fails to amaze and delight me that I can sail with absolutely no sag in the upper panels, but could easily induce it, in precisely the amount, if I wanted to. Even negative twist is a possibility. I would be pleased to learn from people who know more than me about the pros and cons of "easing the top panels".
........................
Some further thoughts:
It seems to be a "given" that camber should be reduced in the upper panels - for very good reason (the upper panels being the "storm sail", and in strong winds less camber is preferable to more). The Western junk sail originally evolved from the need to provide a safe, easy handled rig which would be suitable for single-handed ocean sailing and capable of encountering gales at sea without the need for fore-deck work and sail changing. I would not dispute the wisdom of all that, though I do question it as a general rule that must apply to everyone. Mauro has raised the very interesting idea of having fully cambered upper panels and I dare say this is something we have all thought about from time to time.
Greatly over-canvassing a boat (on the "given" that a junk sail is easy to reef) is a good idea, up to a point, but embodies a law of diminishing returns. A reefed vessel drags a certain amount of naked mast, and there reaches a point where the drive of a large, heavily reefed sail, when trying to drive to windward, would not overcome the drag of the associated large amount of naked mast.
For inshore sailing, perhaps an argument could be made for having moderate sail area, and upper panels that provide lots of drive, even in strong winds.
I guess this business of sail area, and distribution of camber, is a matter of compromise and harmony, and what is optimal is very much a matter of suiting the intended cruising grounds.