To be honest I'm not totally sure how it works in terms of accessing the patent info, procedures etc. Though I'm fairly sure it costs to look at a patent even if you don't want to license the procedure which is a bit shit.
You only tend to be able to find out what the patent is for in terms of the end product, not how they got there which is obviously the important bit and more often than not the bit which is actually patented. Though you can patent composition of matter as well as method etc
I think there is more of a trend within Universities to go down the patent route with new discoveries/developments rather than open publication, whereas in the past that was more of an industry/business thing to do. No surprise really due to the lack of funding these days so if licensing patents out is a way to make the University money it makes sense. Personally it's hard to gauge which is better.
Patenting does, without a doubt, restrict scientific development.
There's been a number of patents that have been filed for genuinely useful/revolutionary products, one being carbon nanotubes, where the patent owners were charging far too much to license the patent and were only licensing to certain (and not many) companies so no advances were made in the area until the patent expired after about 10 years and the product became available for anyone to do anything with.