Read other letters about this article
First, science journals are not old and stodgy. The two biggest in the field, Nature & Science, are actually more like magazines than journals & will turn down research not because it isn't important, but because it isn't sexy enough or they published too many articles on that topic lately. If you find them difficult to get through, you probably just haven't spent enough time picking up current vocabulary and concepts.
There are of course stodgy journals -- and also ones so radical they are full of crap & die within a few years. There are also many excellent journals well worth reading.
Second, lots of journals make money, and some publishers (notably Elsevier) make a fortune by buying up the "must have" journals for a field then cranking up the subscription rates. The American Psychological Association produces the Journals of Experimental Psychology & uses the proceeds to support themselves --- and they cost very little to subscribe to. So I can't believe it costs Elsevier $1000 a year per subscriber to publish its leading journals.
I think it's fine for journals to charge reasonable rates to pay their publishers & maintain their service. It is useful to have a centralized organization for peer review (though notice the reviewers and most of the editors are not paid!!) and a regular index of good articles. But I also think scientists have to be allowed to publish their articles in electronic form on preprint archives, on their web pages, etc. Nearly all science is paid for with public money, and deservingly so, since open access to knowledge helps boost our economy.