I absolutely agree with your assessment that, if you own the database, you own the illegal content. I don't see two ways about it. You're at risk.
However, as a stopgap measure you should use the burden of proof against any accuser and make it at least difficult for them to obtain proof. No proof, no warrant, no indictment e.t.c..
Now, I'm not versed in the technology, but a Steem node is a server, right? And each witness controls their own server then, correct? So if all witnesses were to run a kind of coordinated blacklist they could prevent any retrieval of specific blocks, before they reach any front end, where, as you point out, some control measures take place, but too late.
In a next step, the database and the witness software will have to become one. This will ensure that the database is useless, unless run through the official witness server software (with the blacklist). I guess there would have to be a hardfork for that, to destroy illegal content up to that point. Also, the blacklist would have to become immutable - once on it, nothing can get off. With such a change, you might still hold the illegal content, but there is no way to get to it anymore.
Legally it's kinda like deleting stuff on a hard drive I suppose. Unless wiped a few times, the data is really still there, but who would go for an indictment on account of deleted data? How would one prove intent of any malfeasance?
Does this make sense?
Yes, each witness controls their servers, which runs their Steem node. The database and software are integrated into one package called Steemd.
It requires 15 or 16 out of top 20 witnesses to co-ordinate on a hardfork. That's what it'll require essentially to remove content. However, to do this every time there's illegal content seems very messy to me.
There's no issue if a service provider removes the content and co-operates with the authorities to find who posted it. But as a Steem witness, especially as a low ranked reserve witness, I don't have the option of removing the content available to me.
Yes, I understand that from the other comments and agree: too messy. What I'm trying to propose is a stepped change of the entire software package. First a quick fix to show goodwill AND make it harder for the authorities to get proof. Then a permanent filter mechanism at the first point of data retrieval.
To my limited understanding it should be possible - is it? (Technically; never mind consensus issues for now. )
That would be possible on a centralized solution like Reddit. Not practical on Steem.