They are written vague on purpose, because to be specific, would allow others to build upon your patent, and patent their improvements, locking you into a stale old way of building said invention, never able to improve it.
This is precisely the type of abuse, by a handful of unscrupulous assholes (patent holders being, relative to the entirety of the population, a handful of people), which I propose we amend patent laws to prevent.
And, by amend, I truly mean "actually enforce the laws as written", since they already require some degree of specificity.
I am a patent lawyer, and I completely agree. My patents, of course, are clear and informative; but yes, there are many terrible ones out there. Frankly, it's partly unscrupulous assholes, but mostly incompetent and lazy assholes: to write a good patent application, you have to understand the invention... too many patent lawyers skip that step, take whatever the inventor sent them and slap some boilerplate "in some embodiments" language on it, and file it. Heck, you can still charge the same amount as a well-written patent, but can crank it out in an afternoon! What a world!
Fortunately, the courts and the patent office are finally pushing back on this. Most of the "abstract idea" rejections under Bilski and Alice Corp and other related 35 USC 101 cases are really about badly written patents that claim "A method for doing something awesome, comprising: applying rules, by an expert computer system, to do something awesome." What rules? How does it achieve that awesome result? Fark if anyone knows... the person drafting the patent sure as hell didn't. The cases that are being upheld are the ones that go into detail about what calculations are being performed, how the thing works, the low-level specifics of what it does, etc.
That said, patent law and courts and such are glacial. It'll be another decade and change before patents drafted and granted, say, 5 years ago, expire. And patent litigation with terrible patents will keep popping up over that time. But maybe by the 2030s, things'll start looking better. \_()_/
It's debatable whether the term should be shortened; many would argue it should be extended, as was done with copyright. Personally, I believe that patents and copyright were given the terms they were originally given based on how long it took to produce and circulate a work at the time that those respective laws were written; as both now take considerably less time, yes, I agree that the terms should be shortened.
Patent term has only ever been extended twice, and the second one wasn't a real extension (the change from 17-years-from-issue to 20-years-from-filing was based on an average 3 year prosecution queue, so the result is the same). Copyright has big money publishers on one side like Sony, Disney, Columbia, etc. wanting longer term and, what, pirates? The public? No money on the other side. So your bought-and-sold Congresscritter happily votes for term extensions.
But in patents, Apple, say, wants longer terms for their own patents, but shorter terms for Google and Microsoft's. And vice versa. So you get this pressure on both sides, with no real imbalance in money and lobbyists.
Incidentally, there's a safety valve in patent term already - patent owners have to pay maintenance fees that increase over the life of the patent, or it goes abandoned. Most patents in the tech sector are abandoned long before that 20 year term expires, because, after 10 years, say, they're obsolete. It's the pharmaceutical people who try to keep them alive until the very end, because of how long R&D and FDA approval takes. Increasing those maintenance fees would have the same effect of shortening patent term in fast moving industries while keeping it long where it's needed.