Normal failure rates expected in conventional weapons is 5-15%. A failure rate of 1% is an outstanding achievement. The RAF have only withdrawn the Mk 1 BL 755, Mk 2 is still in service. RAF have also used BLU 97 which is the US system. Counter battery fire using bomblet requires significantly less guns than conventional 155/105mm HE Projectiles to achieve the same effect. Cluster munitions are therefore a considerable force multiplier.
It's all well and good banning the damn things but there are two things that you have to take into account.....
With the forces getting ever smaller, there is a need for these 'force multiplication ' weapons. If you ake something out of your arsnel you should replace it with something that does the same job or better.
We may be moved into banning them, but there are plenty of less 'careing' nations that aren't going to bother..... so what's the point other than making us weaker?
There are lots of variables when it comes to failure rates, not least is the kind of ground you're dropping onto. Trees allow the munitions to decellerate, the ones that come down via parachute often get snagged (imagine the target that presents for an 8yr old with a case of shiney kit syndrome) and soft ground (Iraq, Laos) means that failure rates can be as high as 30%.
You're right about having to factor in the number of munitions though. Let take Laos for example. From 1964 to 1973, the equivalent of a B52's worth of ordnance was dropped on the country every 8 minutes and the UN estimates that there are 500,000 Tonnes of UXO lurking out there. The ICBL has estimated from US procurement figures that 90 million BLU 26 submunitions were expended. 33 years after the end of the war, 25% of villages in Laos have their environs contaminated with UXO- and when you're a subsistence farmer, not venturing out into the fields is not an option.
Now then, let's take a look at how effective they are. The US Munitions Effects Assessment Team reached a vastly different conclusion regarding CBUs effectiveness in Op Allied Force:
The JCS released a report, based on aircrew interviews, that suggested that 140 Tanks, 220 APCs and 450 artillery pieces were destroyed using cluster munitions.
The MEA Team concluded that cluster munitions had destroyed just 14 tanks, 18 APCs and 20 artillery pieces and of 744 confirmed strikes conducted by NATO using CBUs, only 58 were successful.
So, in addition to the jus in bello arguments regarding proportionality, risks to CIVPOP etc., there are good, pragmatic reasons for getting shot of them- namely they don't fecking work!