I research various topics on a regular basis. My first preference is a recent (within, say, the last five years) guidance document or white paper from a federal government agency or research lab. Next would be technical guidance from the state the project is in, or a state in the same region.
The problem is, though, that if it's a research project that I'm working on (as opposed to an intern or entry level scientist), I'm usually in much deeper and am working with more technical details: which equation would be better for this application? Are there any specific chemical/physical/biological reactions that I need to figure out? That's when I need to start trawling through the journal articles. And most of the time, the sorts of details I'm looking for are in the meat of the article and aren't listed in the abstract.
There has to be a happy medium between open access (free!) and paying $35 for an article that I don't even know will be useful until I've paid for it. I can get a few articles here and there that have been posted by the authors or are actually open access. Some of my colleagues have memberships that come with journal access and they can send me stuff. But I can't justify the cost of spending a couple of hundred bucks to trace a possible dead end.
In grad school, our print shop had an arrangements with the publishers that they would copy journal articles, charge us a reasonable price (I think it was a buck or so a page), and send on the royalties as appropriate. I don't get why the publishers can't charge a more reasonable price (say, $5) for a reasonably short article. I'd be able to actually pay for quite a bit more if I could do so in smaller increments.