Depends on the situation. Are you talking about mocking up for a feature where you are recording the score or just doing mock ups of existing pieces?Chris, I'm curious about how you'd typically handle mocking up a string part on CSS if the score calls for divisi. Say the cellos split into three legato parts for a bar, would you just use the CSS cellos three times, or maybe switch to SCS or Afflatus?
For the gigs I've done. We rarely use divisi, mainly because the type of projects I've worked on benefit from just traditional scoring within the one string section = one part mentality. If we do divisi, we tend to just turn off legato and sequence the lines polyphonic-old school style. Since I layer with other libraries, you can hide the seems very well. The orchestrators that I've worked with also make their creative decisions. In many cases, they avoid divisi, especially if we have a small recording budget and can't get a large string section. Overdub limitations make that a challenge. So, I've learned to write more simplistically whilst not trying to be creatively limited. Either way, having a good orchestrator makes a difference. Good ones take what you've done that sounds great with samples and alters it to sound great for the live players you get to record.
If you are talking more about doing covers of existing scores. It depends on the library. For something like CSS, again, I just turn off legato and write polyphonic arco lines in a single patch. CSS has a small enough string section that you can get away with it. The same goes for other libraries like Spitfire Chamber Strings. I've also used Cinematic Strings 2 as a "B section" to CSS to write divisi harmonies with a lot of success but again, I rarely write divisi so it's not a huge concern for me.