Wikipedia entries are not the end, especially if you are doing academic research. When I read undergraduate research papers, a wikipedia entry (or worse, if the whole paper is cut-and-paste from wikipedia!), as sole reference would most assuredly draw my ire. Many undergrads (and a few graduate students!) who use wikipedia probably do not understand the limitations of wikipedia.
There is a nice entry in wiki itself on how to use wikipedia for research, but it's too many clicks away from the main page. To get there from the main page, I clicked on help, and then on Browsing Wikipedia, after which, I then clicked on Doing Research Using Wikipedia. Most regular users, I suspect, would skip the help pages. After all, you just plug in the search box, and then wikipedia will give a list of entries that may be relevant.
One of the reasons to be cautious when using wikipedia as a primary reference, is, paradoxically, because it is an encyclopaedia. Encyclopaedias are not meant to be the last word; they may act as a quick reference, but the meat is in the original or primary sources. A good scholar will track the primary works because the primary sources will be much more careful than secondary sources (of which Encyclopaedias are an example). Primary sources are much more likely to explain the limitations of the study, and will be less likely to contain outrageous claims.
Even within research articles, there is a tendency to be lazy when it comes to citing primary sources. I remember reading an editorial in Nature Cell Biology 11, 1 (2009) where the editors asked that authors be more careful with citations. Sometimes, researchers will just copy and paste the references used in papers they cite, even though the references themselves may be irrelevant, and then because of page-number limitations, forget to cite the true pioneering work.
Second, Wikipedia is editable by anyone with online access. Sometimes, there will be people with sinister agenda who re-edit pieces to suit their beliefs. This means garbage can be created at-will, and some time may pass before another venturesome soul will decide to create a more balanced account. After that, there may be some back-and-forth editing before a consensus view emerges. Even then, we cannot really be sure that the consensus view is correct. This will inevitably bring us back to a search for primary sources.
Third, there will be hoaxes. One of the best ones I know of is the Brahmanical See, a former wikipedia entry in which it is claimed that Hinduism has its own equivalent of the Pope. This entry lasted for approximately three and a half years before it was caught by a zealous editor who could find no supporting documents. Which brings us back to to the gospel of looking for primary sources.
There are a lot more reasons (of which I am too lazy to state), so I'm linking (ironically) to two entries in wikipedia: Why Wikipedia is Not-So-Great, and Wikipedia is Failing.
Thus, one should look for the primary sources. After that, one needs to know if the primary sources can be trusted. Arnold Arons, a physicist I admire, has always emphasized the questions "How do we know?" and "Why do We Believe?", and these questions should always be your company when sifting through primary sources. These questions, by the way, form the foundations of epistemology, and answering those questions, in turn, is surely a story for meant for another day.