When I first started at ProdigalU, I had a pretty good sense of where I should send my work to get it published quickly (or at least reviewed), primarily from my experience at National Lab. We had internal review there, so I was involved in many more manuscripts than just my own work. Now, though, I find that it is hard to say what will go in quickly, and what will need to be shopped around. Definitely part of it is the huge rise in the number of submitted manuscripts. There are so many submissions that an editor has to process each one quickly. As more and more countries improve their science infrastructure, this will only get worse. A complicating issue is the huge increase in the number of journals, particularly specialty journals muddling the scope of older, more familiar names.
I actually don't care much about the name of the journals I publish in (especially now that I have tenure), as long as the journal is indexed in the most important databases. I find that my work is still cited at the around the same rates, even when I publish in lesser known journals, with the exception of journals with extremely high impact (like the Nature babies). However, my students are in the stage of their careers where names on CVs matter, so I still contribute to the problem of "impact hunting", and submit my manuscripts to the most "prestigious" journals that might actually accept it. Sometimes, I find myself submitting along a chain of 3 or 4 journals before review. Most are quick (a few days), so there isn't that much of a delay, but it is inefficient, and demoralizing to the student.
I tell my students that rejection prior to review is just one person's opinion, and that even after review, it is just 2 or 3 people's opinions. I remind them that my most highly cited manuscript was rejected 3 times prior to publication. When we are preparing a manuscript, I make sure they understand that the target audience is more important than the impact factor, that publication in a society journal is often the fastest and most appropriate way to get their work seen by others, and that citations are more important then impact factor. I also make sure they know that if they plan to stay in science, they better get used to rejection.
One of the issues in targeting a manuscript with all the new journals is in looking to see if our work will fit. There are so many journals, that even the best library can't possibly subscribe to them all. We tend to stick to journals ProdigalU subscribes to, even though I never really have problems getting any particular paper (whether we subscribe or not). The problem comes when I want to look at a few issues to check out how the scope plays out in practice, or to get a sense of the style.
I don't have much experience with open access journals. It is not common to publish in such journals in my field, and I can't really afford to pay thousands of dollars in publication fees anyway. Most journals may as well be "open access" with the rise of SciHub and other methods to get copies of papers. To be honest, off-campus access through ProdigalU is so annoying that I often just Google the title, and can usually find an accessible pdf somewhere. James Heathers has a great post about just this topic. I don't think the current publication model is sustainable, and all these ruminations are part of the reason why.