One of the main issues, as I see it, is that the thresholds for obviousness and prior art are woefully lax. If patents were only applied to "expensive R&D that

One of the main issues, as I see it, is that the thresholds for obviousness and prior art are woefully lax. If patents were only applied to "expensive R&D that leads to a simple implementation", I think there would be much less argument over their benefit to society. But what seems to frequently happen is that patents are issued covering abstract, conceptual ideas which are just rearrangements of existing techniques in a marginally novel way, and which may never have actually been proven by a concrete implementation. That's not really rewarding research effort, it's just rewarding whoever was first to stake their claim and pay an attorney to draft the language convincingly. And what's particularly galling is when a company obtains large numbers of patents like that, but only has the focus and/or resources to turn a tiny fraction of them into real products. So whatever consumer value might theoretically have been created by the others is squandered until the patent expires.