Each scientific manuscript submitted for publication to specialized journals, as well as each proposal requesting research funding from government agencies, is anonymously reviewed by a number of scientists. These individual are carefully selected by a journal editor or a program director, with the goal of providing a competent, thorough and detailed assessment of the merits and weaknesses of the submission. Usually, reviewers are themselves active investigators in the field of research where the subject of the manuscript or proposal falls.
In an ideal world, reviewers would examine the manuscript or proposal carefully and objectively, point out technical flaws, warn the authors of potential pitfalls, mention previous relevant work by others that authors may have overlooked, but also acknowledge and highlight any novel idea or result (either achieved or imminent) contained in the work reviewed.
But of course the world is not ideal, and reviewers almost inevitably find themselves in what is best defined as a conflict of interest. For, reviewers are invariably competing with the authors of the manuscript or proposal, for glory and/or funding. It is probably too naive to ask of flawed human beings not to contaminate a review with even a tiny bit of their own bias, ego, prejudice. But sometimes that “bit” is not so “tiny”, i.e., the reviewer takes advantage of anonymity to try and shoot down the activity of competitors (in the process promoting his/her own work or agenda), with arguments that can often be scarcely scientific, specious, disingenuous. For simplicity, I shall henceforth refer to this sort of reviewer commentary as malarkey.
It would be in principle the job of an Editor or a Program Director to tell apart legitimate, precisely articulated concerns from malarkey, i.e., subtle (and not-so-subtle) attempts at self-promotion or unfair competition . But as we all know, that is difficult and time-consuming, and as a result reviews are forwarded, unedited and uncensored, to authors, who have to respond or, in any case, live with them.
My favorite piece of reviewer malarkey is the so-called beaten to death (B2D) argument. The standard B2D review plays out as follows: the reviewer will attempt to downplay, or dismiss the importance of a piece of work carried out by competitors, or the potential impact of a proposed research, by alleging that the subject has been “beaten to death”, that is, a lot of previous work has attempted to address, entirely or in part, the same scientific issues of interest to the authors.
Such a review will typically be accompanied by the reviewer’s exhortation that the authors “shift their focus”, abandon their current line or subject of research, and perhaps re-direct their effort toward another problem (which may well happen to be that of interest of the reviewer, incidentally).
Obviously, such an objection would have merit if, say, results presented by authors had already been obtained (and published) by others, or if a proposed investigative approach had already been tried out by others, or if the scientific question that the authors addressed (or, intended to address), could be regarded as already answered to a satisfactory degree, if the authors focused on small, relatively minor details. While there would certainly be a degree of subjectivity in such a judgment, serious reviewers could state their case quantitatively by providing references to all previous, cogent work. If overlap with the authors’ content should indeed appear significant, one could conceivably understand, if not always support, a recommendation that authors broaden their research scope.
But in the B2D scenario, a reviewer will not do any of that. No reference will be provided, and often times not even a general statement will be made, summarizing the status of the research on that subject and contending that the work carried out or proposed does/would not add much of significance .
A B2D reviewer will simply express annoyance and weariness with a subject on which, according to him/her, too much work has already been done, regardless of whether all that work has led anywhere, of the status of research on that subject and/or whether outstanding scientific issues still exist. 
Now, one would think that, the mere fact that authors wish to attempt to solve a problem that pre-dates them ought not, in and of itself, demean the (potential) value of their work, if the question that they pose is still unanswered. In fact, that a problem has remained unsolved for many years is, if anything, an indication that something potentially very important could be discovered, and therefore further investigation (obviously based on sound methodologies and original ideas) would only be welcomed — since when do scientists quit working on problems because they are too difficult ? .
Yet this is not the opinion of the B2D reviewers, who often cite the very fact that all previous work has failed as a reason to abandon that problem altogether. They will basically tell the authors that because that particular question is no longer high on their priority list, no longer excites or entertains them, everyone else should be discouraged or prevented from tackling it.
On occasion, B2D reviewers will actually ignore altogether possibly significant new results presented on a manuscript, even if they unmistakably contradict the “conventional wisdom” on that subject, simply because “that problem has been beaten to death”. And yes, incredibly Editors and Program Directors go along with such nonsense.
To some extent, again, this is perhaps inevitable. For all of its pretense of objectivity, the scientific enterprise is affected by the same dynamics observed in other human activities. There are fads, of course, and at a given time the community will invariably feel that some subjects (e.g., those suggested by a few revered researchers operating at prominent institutions … just kidding) are more important or timely than others. Sometimes a reviewer may want, in good faith, encourage talented colleagues to work on subjects where they could make a bigger “splash”, eliciting the attention of a greater part of the community.
Of course, one could also find the occasional, lazy reviewer who does not want to bother to read the literature and actually understand in detail what others have done or are doing.
Unfortunately though, often times I am afraid that such shallow, groundless and non-scientific reviews underlie individual agendas. But, why would one write, e.g., on a proposal review, something like “I do not think that you, the author, should work on subject A, for it has been B2D… work on subject B instead” ?
I can think if at least two self-serving motives behind a B2D report:
Anyone can think of others ?
 As a case in point, I never cease to be amazed by the number of journal Editors who regard it as appropriate of reviewers (or, referees, as we call them in physics) to claim that prior work had already obtained some or all of the results presented in a submitted manuscript, without providing a single reference to such work. This is one of those cases where the authors, upon receiving the reviews, will get angry at the reviewer when the target of their ire should really be the Editor.
 Obviously merely stating anything like that without substantiation (yes, those references again), is insufficient, but still better than nothing.
 In condensed matter physics (at least in America) such is the situation with high temperature superconductivity. Any researcher wanting to seek funding for research on this subject better be ready to be read B2D malarkey all over the reports (it is worth reminding that, over twenty years after its discovery, the physical mechanism underlying high-temperature superconductivity is yet unknown).