No copyright isn't the solution. But it probably would be better than the current situation. Bring back either a 20-year copyright or an "author's death plus 10 years" version. The latter would be easiest to litigate and would incentivize relatives pulling together and sharing papers and unfinished manuscripts.
A good question, though we're currently living in a world that is the author's death plus 95 years in the US, so this is already standard practice. Works not covered this way get a ridiculous 120 years copyright coverage. I would say a flat 35 years from publication or 50 years for unpublished works would be reasonable.
This encourages gaming the system: companies will seek to designate as the copyright holder the youngest available employee, to make the term last longer.
20 year is ok, but "author's death plus 10 years" is bad because it short enough to create some dreadful incentive for would-be publishers: if author is not willing to deal with you just kill him and wait a decade
That doesn't make sense. It's not like the person who killed them would get a copyright monopoly on the work. Literally anyone could publish it. Do you see anyone getting rich publishing works in the public domain? If that's the case, I need to get myself some of that red-hot Project Gutenberg money.