What I've seen happen is that people advance (too) quickly to introductory tech/cave levels on OC, get certified by sloppy instructors, struggle with one or other skill, get disappointed when there's no instant gratification, try to to fix this by getting more gear and finally ending up with an RB, all the while doing dives that could be very easily done with just a little bit of extra training and experience on OC. Seen this locally multiple times, and luckily this far it seems to have ended with the individual eventually dropping out of diving or getting a good scare and realizing it's not about the gear.
Quite often people seem to be blaming RBs for allowing divers to push themselves further than their skills would merit. I'm not entirely sure if I agree, since RBs can kill you just as easily in the shallows - if your fundamental skills (buoyancy, trim, finning, just-f'in-staying-still) are sub-par, you're not going to able to monitor the RB, other dive parameters or your buddy, and are much more liable to get in trouble with CO2. Looking at the "congratulations-for-passing" posts and associated videos at RBW, for example, is an eye-opener. Only a few years ago (at least locally) RBs were seen as a tool for the very advanced diver, now you find folks telling aspiring tech divers to jump in rather sooner than later, and just about every muppet with OWD level buoyancy control and enough cash to buy the device get only a back-patting for "joining the dark side" instead of the "are-you-really-sure" questions you'd get if you, for example, announced needing to get to OC advanced trimix ASAP with just 50 dives.
For the record, I don't dive CC, but I've taken a long hard look at them - the logistics advantages for deep wreck diving trips would be substantial. I'm simply not ready to take the leap with any of the units commonly used. While I agree with Steve that simply adding more sensors etc to the current RBs is not the solution, I'd also like to point out that at the moment, just about every RB is built rather than manufactured (in a non-flattering way), and the loop is essentially an unstable feedback system kept alive by extremely unreliable sensors giving input to dodgy control systems (be it the diver or electronics). It really does take training and cojones to dive one now - the case would be completely different if we had 100% reliable O2 (and CO2) sensors.
As an aside, I have absolutely no idea why many builders seem to think it's a good idea to add deco calculation to an already complicated real-time O2 controller, when the deco could just as well be calculated completely isolated from the loop, and more importantly, rest of the electronics. You don't need to get real-time ppO2 reading for good-enough deco calc, yet customers and manufacturers seem to insist on it. Just look at the problems people are having with Visions, X1s, Shearwaters or OSTCs; while such error rate would be borderline acceptable on a deco computer used by an experienced diver, there's no way I'd want to see similar on an O2 controller. Don't really know if hanging, rebooting or buggy electronics have been to blame in accidents, but it simply does not instill much confidence that the manufacturers really know what's best for me... Slightly off-topic, but needed to vent because I work with complex software (non-embedded, non-real-time, though).
Then again, I have no problem diving with RB divers if they got their stuff together, bring enough OC gas and follow team deco. I'm just too much of a chicken to dive one myself, even if that means blowing between 500-1000eur/year on helium.
//LN