At a presentation I attended given by the Australian Cirrus agents some months ago it was explicitly stated the Cirrus can recover from a spin without using the CAPS, but only if there is sufficient altitude.
That's the key thing - sufficient altitude. In the case of the Blue Mountains accident, there wasn't sufficient altitude, and the only way to save the situation was to fire the 'chute.
Result? Three people walked away alive, rather than being carried away, dead and charred, in body-bags.
The lack of altitude in this case was no fault of the Cirrus aircraft.
We were also told at the presentation that in training the Cirrus instructors use the simulator to put potential Cirrus purchasers (many of whom are highly-experienced pilots) into situations from which recovery is not possible, and these highly experienced pilots repeatedly wind themselves and the simulator into the ground, rather than firing the 'chute.
The habit of flying the aircraft "as far into the crash as possible" (attributed to Bob Hoover) is so deeply ingrained that the idea of reaching up for the firing-handle and pulling it just doesn't seem to occur...until the pilot has crashed the sim a few times, and realises that the 'chute would have saved them. When the sim is put into an irrecoverable situation and the instructor sees the student's hand reaching for the firing-handle, then the message is starting to sink in.
The Cirrus requires a major adjustment to pilot mind-set. It is a matter of re-educating these pilots to use the aircraft parachute system before the situation becomes so bad that even the 'chute is not going to save them because it is being operated outside its design limits. If used within its design limits, the CAPS does save lives, and Cirrus has the statistics to prove this. Go ahead, ring them up and ask them. Ask them how many people have survived parachute deployments on the Cirrus when the 'chute was operated within its operational envelope. Ask them what their injuries were for those incidents.
Is there anything fundamentally wrong with the Cirrus? Probably not. Is there anything fundamentally wrong with many of the pilots who fly it? Quite probably, yes there is.
It is not for the low-hour GA pilot, for a start. It is not a "seat-of-the-pants" aircraft. It has to be flown "by the numbers", the same as any similar high-performance single-engine aircraft. Like any other IFR-equipped aircraft it requires a pilot to be current and up-to-date on IFR techniques to be successfully and safely operated in IFR flight.
Some aircraft are designed to be deliberately stalled, and some are not. I personally think the de-emphasis on stall and spin-recovery training is not a good thing in GA. Even so, there are many GA training aircraft with adverse stall and spin characteristics (Chipmunk and Tomahawk, for example); aircraft which you did not intentionally get into a stalled or spinning situation at low altitude.
The debate about whether it is better to have a training aircraft with a benign stall or a sudden onset with rapid wing-drop has been raging for decades now, each side arguing the their case with the ferocity of wizened clerics arguing an abstruse point of canon-law. However, the Cirrus (or Lancair, or Columbia, etc) are not training aircraft, and therefore not designed to be intentionally stalled or spun.
Equipping the Cirrus with CAPS was not an admission of failure on the part of the aircraft. It was an attempt to give pilots who make a series of successively greater mistakes resulting in loss of control one last chance to save themselves and their passengers.
It is one of relatively few aircraft to do so. It has saved lives in the past which would have been lost, and it will continue to do so in the future. And, even though it may not save the life of a Cirrus pilot who has a fatal heart-attack in flight, if his or her passengers fire the 'chute, it will almost certainly save them.