The advent of open access (OA) publishing has lead to a proliferation of journals which offer a peer reviewed publication venue for a nominal charge.  Some of these journals are associated with scholarly associations.  Such as Physical Review X.  Others are not, such as a journal I published in called the International Journal of Astronomy and Astrophysics (SCRIP) or the journals published by Elsevier, and other such companies. 

However, with author pays open access comes the danger that a journal will have a financial motive to publish whatever nonsense it receives without any pre or post publication review or comments.  As well as no protocol for making corrections or having conversations about a paper which can and in some cased has, ie. The Winnower and Science Open Research, taken the place of actual peer review.  Journals that don't have either organized post publication review or traditional pre publication review are not as reliable as those that do.  Responsible readers should never take the word of a published article in either format as sacrosanct weather it has the imprimatur of a large scholarly body or a small new publisher.   Some due and that is IMHO mentally lazy. 

Jeffery Beall has done overall admirable work collecting a list of suspect publishers.  The problem with his work is it covers such a broad swath of publishing topics and publishers.  While many cases are cut and dried others are not.  Many of the journals have one rather poorly designed website, no peer review, and offer a "discount" for publishing right away.  They spam people etc.   Some top journals do that too.  I was once asked to submit something to the journal Classical and Quantum Gravity, as I recall,  associated with the IOP in Britain.  They did not publish it, but even top well respected journals do that sometimes. 

This is my comment to Jeffery Beall on his work and why he should be a bit more careful about tarring and feathering and setting on fire a whole publisher because of either growing pains, or a couple of bad articles. 

I really think you should reconsider what you have said about SCRIP and it's journals.  They provide access to a peer reviewed, open access, publishing process for not a great deal of money.    I will note I am one who has published work there. 
Farmer, H. (2014) Disk-Outflow Models as Applied to High Mass Star Forming Regions through Methanol and Water Maser Observations. International Journal of Astronomy and Astrophysics, 4, 571-597. doi: 10.4236/ijaa.2014.44053.  arXiv:1311.3983  2014IJAA....4..571F With online data soon to appear in Vizier  An online database maintained by astronomers who have no problem with the SCRIP journal  "The International Journal of Astronomy and Astrophysics".   Work for which I was granted the chance to speak at last Aprils APS meeting which I was unable to attend due to finances and time conflicts.
In fact when I was looking to publish this paper based on my MS thesis I first sent it to as I recall it was either MNRAS or A&A.  The editor there suggested that I publish it in a journal like IJAA.  Stating to the effect, they only publish top tier new observations.  So if you aren't working with the VLA, VLBA, ATCA, Hubble, or eventually the James Webb Space Telescope you aren't getting into the top journals. 

While they do publish some really bad articles sometimes so has Physical Review, so has Nature, So has Science.    Physical Review famously rejected the original paper by Maiman describing the laser (Which he called an optical maser since masers were discovered first.)   He then took it to Nature who published it.   To the person above who said that taking a paper to a different journal to try again there is somehow a bad practice really should consider the power not doing that gives one or two APS editors.  People I know and respect but they aren't Gods. 

It would be more fair to call out individual bad or fishy articles.  Then question those articles. Questioning a whole publisher and saying really caustic things about the journals, editors, and scholars who publish there, primarily because of the low cost* is at best inflammatory and prejudicial.  Especially when some of those journals are acceptable references of record to the scholarly communities they wish to serve.   I think astronomers know better than you what Astronomy journals are fraudulent.  IJAA does not accept everything by a long shot.  My paper was under review for a month, and was in revision for a year.  To get a discount they did let me referee some papers and I personally rejected papers which have not since been published there.   I've seen the inside and it's Kosher. 

*I am considering publishing another paper OA in an Elsevier journal but they want $2500.  So I either go OA somewhere else or subscription there.  Philosophically I don't like subscription based journals.  Better knowledge of the sky should be as free as looking up is.   Would you really think an article is better just because it was published in a more expensive location? 
So who should keep an official look out for truly bad/fraudulent journals? 
Here is my rough proposal  Journals themselves should be periodically peer reviewed by scholarly organizations like say APS, NSF, AAAS, and international bodies.  This would have to be a double or triple blinded process.  Simply having all the editors of journals associated with this or that scholarly organization express their feelings regarding journals other than their own would be dismissable due to natural human bias.   

Randomly selected articles from a given journal will be scrubbed of all identifying information.   Only text should be used.   Figures are too easily recognized.  This would be done by a set of initial reviewers or an automated process.  An automated process would ensure less chance of bias.  

The reviewers should be from the same area of science but not the same subfield.  They should have enough knowledge to know the subject but not so much that they can recognize an author or journal just by the text.  (Such people exist).  

The reviewers should ideally discuss each paper in person but email can do.  Their only task is to determine on a scale of say 1-10 or 1-5 how likely is it that this article is just nonsense based on a set of written criteria.  These criteria should make a distinction between poorly written English and nonsense use of words. 

The reviewers will then make a determination about those papers, and depending on how they judge the papers that would be a judgement of the journal. 

This process could be called a sort of Journal Jury, or more whimsically a Pepsi Vs Coke blind taste test (with RC thrown in too just to check for any bias towards big prestigious names).   This would ensure that no one man, no matter how well intentioned as I am certain Mr Beall is gets to dictate which journal or whole family of journals by a publisher are "valid".