To start with - the usage or public presentation of UDRS technology seems to be more erroneous than the actual technology itself. For example if a fielding team appeals for LBW and the umpire gives it not out and the fielding team refers the decision to UDRS - the first thing that is checked is if the bowler bowled a no-ball or not. If he had indeed bowled a no-ball, we don't even proceed to the next step of seeing whether the ball might have gone to hit the stumps or not. The review process is abandoned right then and there and the the batsman is given not out.
The 2.5 meter rule should ideally be used in a similar gated fashion. Before projecting the direction of the ball - the technology should essentially measure the distance of impact between ball/pad and the stumps. If it is beyond 2.5 meters - it should straight away abandon the referral process and either give 'not out' or at least revert to the on-field umpire's decision.
What the people behind UDRS claim is that the 'hawkeye projection' is more error prone or uncertain when the distance between impact and stumps is greater than 2.5 meters. You may ask why 2.5 and not 2.6 or 3.7? Thats just because you don't know enough. I don't know enough. Maybe the technology manufacturers did several million samples of statistical analysis and found out that until 2.5 meters they have 99% accuracy and at about 2.5 meters the accuracy dropped down to 70%. Who knows? And when you don't know enough the best thing to do is shut up and let the person who knows enough to make a decision. It seems in this case the manufacturers of the technology seem to think that their technology does not work beyond 2.5 meters. It could have been 3.1 but in this case it is not. It has to be some number and in the case of this technology it was 2.5 meters.
So what is really happening is that this technology is opting out of the decision making process if the impact is greater than 2.5 meters. This is a simple enough rule. This essentially means that the ability for technology to support referrals is only within the 2.5 meter scope.
The horrible thing they did in that match is present/use UDRS poorly. So poorly that they had UDRS project the path of the ball even when self-admittedly the projection was inaccurate and unreliable. They did not tell us before showing the visual that projections beyond 2.5 meters was unreliable. The disclaimer was written on the right-hand side in red font that "2.5 meters bla bla bla". Which no one read. Rightly so - because no one ever reads the fine print. The fact remains that - they shouldn't bother projecting/showing us stuff that has no reliability. Might as well tie a pen to a chicken's feet and show that as a 'projection' on the big screen and call it UDRS. Now what that colossal stupidity led to was - millions of people saw some random projection - came to believe that the ball was *certainly* going to hit the middle of the middle stump. And they all got angsty when Billy Bowden gave 'not out'. And Dhoni comes out and says that technology was "adultrated by Human intervention". When in reality "Adultrated Technology removed itself out of the equation and reverted back to human's best decision".
You could ask what should be done under circumstances when the ball hits the pad when the batsman is well in front.
When technology with all its power is unsure at 2.5 meters - I do not think on-field umpires can be sure either. And when there is doubt the batsman should be given not out. So the sensible thing to do is to fix a threshold (be it 2.5 or 2.6 whatever the technology decides as statistically favorable to making a more accurate decision) and set that as boundary condition. Impacts over 2.5 meters should be given 'not out' by both on-field and UDRS. You may ask how an umpire is to know whether the impact was 2.5 meters or more. Well - he uses the same guesstimate he uses to decide whether the ball pitched in-line or not. In cases where he gives out for impact greater than 2.5 meters, the batsman can appeal and get a 'not out' from UDRS. In cases where on-field umpire says not-out and fielding side is unsure - the fielding side can refer to UDRS and if the impact was less than 2.5 meters - UDRS will go on to project the path of the ball and deem it out or not out based on that.
But the key thing to do is not show the public stuff that does not have a high degree of certainty. Then the whole discussion gets biased by what people saw. Visual images can bias a person beyond reason. And it ridiculous that the ICC does not even know this basic fact of life.