Hi xarax,sorry for the extreme lateness of this reply
Luca, I suppose you have tried those three variations, shown again in Reply#5 here[ http://igkt.net/sm/index.php?topic=4821.msg31427#msg31427 ], for some time now. If you would like to evaluate them, what grade, from 1 to 10, would you give to them ?
I prefer the first one, and, if the rope is very sleek, I entangle the Tail ends with one "twist" : the result seems a rock-solid hitch - the only problem is how one unties this thing !
At this point I think you have done your trials, and have derived your own conclusions...
I personally have tried these binders using them around a bulky stack of old newspapers, and around some piles of large diameter with respect of the diameter of the rope used,so I have not yet tested them in a real "mid-air situation".
It is quite difficult,maybe especially as regards the two binders based on the Clove hitch,to make sure that the hitch component remains well compact:the problem is that, by pulling the free ends to tighten the binder around the object, the two "round-turned" components of the Clove hitch tend to stay loose,with consequent loss of time (and possibly also of patience ...) in searching the way to tighten ...
The Prusik version is perhaps a little faster and more intuitive to realize, but the described situation is not much improved(but I have to remember that these three binders are born from a precise request of Kieran
http://igkt.net/sm/index.php?topic=4631.msg29976#msg29976 , so these binders have been designed to be realized only once, and then reused many times without ever having to untie and re-tie them).
In any case, at the time of that thread, to overcome the problem described above, I had thought of entrusting the "bite" around the tail ends exclusively to the round-turned components of the hitch, leaving slack the "bridge" that unites them,and using this same "bridge" by pulling it(unless one decides to use the knot upside down, with the "bridge" which in this case remains in contact with the surface of the wrapped object) in order to take the opportunity to unite and further tighten the two round turns ... do not know to what extent it is a good idea, but apparently it seems to work!
Or (but in this case the surface of the wrapped object must be large enough to allow it) the two round-turned components can be moved away from one another for than allows the length of the "bridge" ,to create two distinct points where the tail ends are nipped.
OK,I realize that I'm not answering your question ... well, the fact is that I can not do it!
The Prusik version seems a little more "easy" to dress and set,and the tail ends form angles of 90 degrees at the time of entering through the hitch component,but I am not able to quantify which of these binders maintains tension better than others, I think that depends a lot on how one succeeds to dress and set them(it's a struggle!) and the materials used.
Using the Prusik binder as described at the beginning of the post, I tighten this binder by pulling both the tails using only the(poor!) force of my arms,without using any sort of mechanical advantage,so I never found it difficult to untie it.
Trying to keep compact the knot's nub of the Prusik component, I think it's very hard to make sure that the two external turns of the Prusik are gripping at the same level of the two interior turns,so I do not know how much it's worth it compared to the use of a simple Girth hitch!
Bye!