Yeah. The whole Section 1201 is just terrible.
And the "innocent violations" thing is burden of proof on the "infringer", and is subject to court discresion. You can prove that you neither knew, nor had no way of knowing that the thing you distributed was designed to defeat a technological measure and still have the court say "too bad, that is the risk you take distributing things, you owe every cent they lost via piracy attributable to this plus statutory damages (despite the fact they already sued other people and got all the money they lost from piracy from them already)".
And the infringer is anyone who "manufacture, import, offer to the public, provide, or otherwise traffic in any technology, product, service, device, component, or part thereof" with respect to an infringing device/program. note the "part thereof". While intended to allow RIAA/MPAA to go after people who are knowingly benefiting from selling some part that people found can be used to construct a piracy device, it technically means they can sue Linus Torvalds if Linux is used in an infringement tool and win if the Judge is not exercising that discretion.
Everything about 1201 et seq is poorly thought out.
I actually don't mind some other parts of the DCMA, like the basic safe harbor concept, but do believe it needs to updated to make false takedown notices actually carry real consequences for the false claimant (including eventually losing the right to file takedown notices without suing and winning first). Furthermore, that should also cover any similar system implemented by platforms, such as false uploads to a ContentID database, or false use of manual claims.
But yeah, there are a lot of changes/reforms that really ought to be made to US copyright law, to better match what people actually expect the law to be.