There are no oracles of community opinion. Sometimes an editor will use a phrase along the lines of "the community is calling this into question", or "the community does not want […] " as an attempt to gain the upper hand in a debate. It acts as a rhetorical fallacy, painting any opposing view as "not belonging to the community". This is often seen in interactions with "The Foundation", but nearly as often is directed towards the few prolific editors who actually write Wikipedia.
"Oracling" may occurs when thousands of hours of volunteer time have been spent, spurred by a single individual who is made aware of something and makes it out that there is a sudden "public scrutiny" and that changes have been "snuck in through the back door". This ia equally fallacious reasoning, as there are no back doors to consensus (or at the very least, they are few and far apart).
If something goes unquestioned for years, and is being questioned now—that's what matters. Prior history is not incriminating unless editors were knowingly in violation of policies or community decisions. They almost never are, and implying so without clear evidence violates WP:AGF.
Taking the role of 'interpreter of community opinion' is often used to mask true debate, and it is performed by writing a wall of text that discourages others from getting involved. It is a very effective strategy, and best overcome by ignoring the rambling (and often incoherent) message and pointing out that the user is "oracling".
"Oracling" relies on a great deal of arrogance in assuming one's ability to interpret the "will of the community". The decisions and wills of the community are by definition ephemeral and changing, and no single Wikipedian has any special power to elucidate them. We have some methods to gauge the current state of community opinion, such as Wikipedia:Requests for Comment, talk-pages, and noticeboards. But without clear results run through such systems, interpreting the community's wishes and wants is to stand on shaky ground. It may give way at any time, and like the oracle perched above the abyssal gorge in the accompanying image, you may fall into the depths below when you find your interpretation does not hold. Implied consensus is an important aspect of Wikipedia, and the fallout over an issue on Wikipedia and subsequent debates to determine consensus does not mean that all cases where the losing party reigned were wrong prior to the debate—or that any hypothetical community "would have disapproved had they only known of the issue earlier". How the community decides in a certain debate depends largely on who was there, at what time the question was asked, by whom it was asked and how it was formulated, who voted early and on what option, etc. If you do not believe this intuitively, there is ample evidence in the political science field, including that we vote according to how we perceive others to have voted—which obviously influences the open ballot system that is Requests for Comment (no matter how much we say that polls aren't votes). Therefore such a statement is a clear case of hindsight bias, and it is often very likely that a different result would have been reached, if perhaps only slightly.
This doesn't mean that the result of any given RfC is de facto wrong, as that would be to cast aside the notion of the possibility of achieving anything resembling consensus or democratic choice. What it does mean is that you should be very wary when trying to interpret consensus and the will of the community, especially when there isn't a clear and longstanding consensus. Even in the face of overwhelming consensus you should also take care not to assume that the most extreme interpretation of support is the true consensus. It is also important to remember that consensus changes, and so does "the will of the community".
No single Wikipedian is entitled to "interpret the will of the community" or act as arbiter of "community consensus". Experience may make you better at gauging what the community is likely to decide, but no one is ever certain—and Wikipedia's non-hierarchic nature demands we lend an ear to newcomers and those of dissenting views, just as much as we should listen to entrenched users. When two or more sides exist in an argument on Wikipedia, those subscribing to the majority viewpoint should never categorically ignore the minority one.