Now this is going to get some peoples blood flowing...when they first reported on their failure rates of RROD...they are slammed by M$...I'm sure M$ is going to love what they have to say now...even if M$'s official line is NO COMMENT:

http://kotaku.com/5350590/report-360-more-reliable-repairs-less-frequent

According to data released by third-party warranty providers SquareTrade, since the mid-point of 2008 the Xbox 360's reliability has improved out of sight.

Now, before we go any further, know this: we have our doubts when it comes to SquareTrade. A third-party warranty provider? Who nobody had ever heard of until they released some selective statistics? This isn't exactly hard science.

Then again, when Microsoft (or major retailers, for that matter) refuse to disclose official figures on the subject, outlets like SquareTrade are all we've got. So take these figures with a grain of salt and jump on in.

Interestingly, to go with the 360 figures, SquareTrade also said that while 23.7% of their 360-owning customer base reported a system failure, only 10% of PS3 users and a miniscule 2.7% of Wii users had any problems.

Truly upsetting for people who actually believed the 52% fail rate. I think its safe to say Squaretrade has a little more validity.

That's one problem down ...

They are only reporting on their warranty systems...unlike the Game Informer survey, which wasn't scientific, but instead statistical.  The 2 really can't be compared directly, but you can extrapolate some trends from it...like the Jasper chipset is actually doing what it was design to do...and what deliver the console that M$ SHOULD have delivered on launch day.

I'm sure that recent models are more reliable, but isn't it obvious that the chart would point down? After all a machine bought in Q109 is only about 6 months old.
Shouldn't the comparison be something like "percentage of RROD after 3/6/12 months of use" to make proper sense?

Update: I just saw the vertical label stating that this is a "12 months of use" chart indeed

The methodology you are suggesting leave far too much leeway to manipulation of the data, and there are far too many variables involved with you using those metrics.  The way chosen presents raw data, and you can extrapolate a general trend, based on it, especially given all the unknowns.  Some problems with your methodology:

1. This amount of 360s involved are not in a controlled group, so let's say someone bought there 360 several people bought their 360s 12 months ago, and only played it scarcely, while somebody bought their's 3 months ago, and logged more hours than they did.

2. What was the titled released during that period, and how did that impact failure rates?

3. How did updates and features, like NXE, Netflix, etc effect failure rates?

4. What and how is your Xbox configured? ie...laptop cooling device? Intercoolers? Space ventilation? How clean is your power source? 

These are just some unknown variable that make it worthless to use a 3/6/12 month usage chart, since you DON'T know the usage, and the only way to have used your method, would have been for someone to buy a large sample say of 360s, set up a "sterile" lab environment, and measure the data.  When these guys get a Xbox in for repairs, they know 2 things for certain...the manufactured date, and the date it was sold...trying to model usage based on that is terrible engineering analysis to determine any kind of MTBF for the console.

Wow the Wii is super reliable Go Nintendo!!!!

For Nintendo...simpler was better...they re-cycled the most from the previous gen and it worked like a charm.  Sony and M$ when on the hunt for the techinccal holy grail and it cost them in both sales and higher reliability. 

Why? Is this report now become fact. Those who believes the 50% or whatever will carry on doing so and MS fans will obviously take this report as gospel...or at least some of them.

Credit to the OP to state that this report "must be taken with a grain of salt".

I am pretty sure after 4yrs reliablity has improved.