If everyone who sells MSDN accounts has negative trust (and a "WARNING: trade with extreme caution" tag), then it will not come as a surprise when someone who sells MSDN accounts has negative trust. If someone often trades with people who sell MSDN accounts, and the comments on their negative trust always say "this guy is scammin Macrosoft becuz of TOS", which they do not care about, nor do they think is a valid reason to conclude they are a scammer), then it becomes pointless to read the trust comments because they all say the same thing.

The same thing is true for those who participate in ponzs (who run ponzis), who deal in accounts, and who do anything that a select few members simply do not like.

A trust score is necessary to measure how many people have had positive financial interactions with the person whose ability to be trusted is being measured. If only one or two people have had a positive trading experience with someone then, all else being equal, they are less likely to be trustworthy then someone who has had positive trading experiences with 30 people (who are in your trust network). One rationale behind this is that the positive trust rating that the one or two people left might be inaccurate, another rationale is that someone might be willing to walk away from a trade (scamming their trading partner) if something goes wrong, and the one or two people who traded with someone just might have been able to avoid anything from going wrong in their trades.

One other reason why trust ratings are important is because they help people measure differences in trust levels at any point in time. For example, if you decide to trust someone today because you have read someone's trust comments extensively, and otherwise researched and determined that is it safe to trust them, and subsequently have a positive trading experience, then a month from now, if a similar trade were to be proposed to you again, and the person you dealt with all of a sudden has negative trust (when they previously had positive trust), then you might want to look into their ability to be trusted again, but if their trust level has not changed, then it might not be necessary to research their ability to be trusted a second time.