Franco, funny you put that up - I was just looking at it a few days ago. To me, those charts are meaningless. I could not find the basis for these ratings. For these to be meaningful measurements, the rationale behind these ratings needs to be clearly explained and defined for each reliability metric.
They also seem to fluctuate wildly over time and time related variability is not correctly explained (clearly to the reader). I recall looking at a very old issue of this and seeing my box year's rating to be substantially higher. I feel it is very difficult to interpret some of these results
.
Furthermore, (I beleive those ratings are based on frequency of occ.) there are many more factors to consider than mere frequency. For example, cost and cost normalized for MSRP should also be included to further understanding of true reliability problems. After all, isn't everyone's bottom-line here long run expected cost of repairs?