I would not expect toshiba to take action if just one person phoned this in, of course, and I totally agree that to some third party individual anecdotes are not as meaningful as statistical data. But there is statistical data, and then there is statistical data ;). I also doubt that toshiba would take action just because one person phoned in and said they personally had seen this 200 or 1000 times. The value of "statistical evidence" depends upon how it is gathered. Eg, if you took it to court, the fact that you had 200 busted computers would not be enough evidence unless you could provide further evidence that they busted because they were flawed.
Evidence of that sort is about logic and empirical demonstration, not statistics. I think in mathematics there is the concept of proof, and proof is not about having 200 calculators that say 2+2=5.
Identifying a design flaw is similar -- you need to find the flaw, not just the failures that imply a flaw. Statistical failures only imply the presence of a problem, and may help you to find it, but they do not prove that it exists.
So my point was, if I broke a power jack on a computer in near darkness one night stumbling drunk while talking on the phone, I might regard that, as you say, as coincidence. However, if I broke it while paying full attention, being careful, in a sound state of mind, I would not have to do that more than once to recognize this mechanism is excessively delicate. Observing something 200 times does not necessarily make you a better observer.
Statistics are also very prone to abuse and interpretation (because they are not hard evidence). Take your statistics with a grain of salt. CommonTater saying, "I've been in charge of 250 laptops" is a good reason to pay attention if he has an observation about laptops, but it is in no way proof that what he says is accurate, or to put your brain on hold while considering the possibility.
So, in regard to the original topic,
1) the statistical evidence online buries CT's (essentially anecdotal) sample.
2) altho CT's statistics (if accurate) may imply a certain conclusion, a coherent theory and some logic is required to connect the two, and I have not seen that. Just the "argument to authority" stuff, and a lot of rambling. It is simply too easy to come up with any number of explanations why one person would have observed (or believe, or claim to have observed) a phenomenon in a particular context.