> Surely if a previous DOJ refused to prosecute the case it is 'weak' in some sense.
No, prosecutorial decisions are not, even in theory, made based solely on the strength of cases; while that is a factor, evaluation of importance (including cost/benefit considerationa), which are ultimately policy decisions which different decision-makers (and even the same decision-maker at a different time, particularly if facts pertinent to the prioritization but not the strength the of the case change) are likely to see differently even with the same view of strength of the case, are also a factor.
Plus, available evidence and relevant case law can change over time; even if the case was weak during the previous Administration the same case might not be weak now.
> Surely if a previous DOJ refused to prosecute the case it is 'weak' in some sense
That's not a matter of the case being weak, that's a matter of policy on which actors with different policy preferences will differ even with perfect information and judgement regarding application of those preferences.
> Or perhaps because there is prima facie evidence of a crime but First Amendment arguments could potentially prevent the government from securing a conviction.
That would be weakness, but no one has made a coherent First Amendment argument that would bar prosecution for conspiracy to break into a government computer system manifest in an offer to help break a password and actual attempts at that. A lot of emotional appeals lacking a specific argument have been made in that direction, but that's not the same thing.