The perjury clause doesn't say what you think it says. If I own the rights on work A, to file a notice on work B, I claim that work B infringes work A. The perjury clause kicks in only if I do not own the rights to work A (or represent the person who does). If work B doesn't infringe, then that's a matter for the courts. This is quite annoying, but it does make sense. It's clear cut if works A and B are the same, but not in the case that B is a derived work of A. A court has to decide whether the use of A in B counts under fair use or not.
The counterbalance for this is that the DMCA does indemnify YouTube if they respond to a counternotice and reinstate the work. If you, the owner of work B, think it does not infringe then you send such a notice to YouTube. I then have no further recourse against YouTube and must take you to court directly.
The problem here is that it's very easy to automate sending takedown notices, but very hard to automate sending counter-notices. Mass-sending of automated takedown notices was something that the authors of the DMCA didn't foresee and the act probably needs amending to require the notice to explicitly state (under penalty of perjury) the person who has compared the works and their reason for believing that they are infringing.
I feel it necessary to point out, though, that OS X is not a microkernel system comparable to Minix
While this is true, it's worth noting that a lot of the compartmentalisation and sandboxing ideas that most of the userland programs on OS X employ (either directly or via standard APIs) have roots in microkernel research. OS X is in the somewhat odd situation of having userspace processes that are a lot more like multiserver microkernels than its kernel...