It's really hard to see what the issue is without being able to reproduce it. Those times seem unusual to me, changing the distribution maps shouldn't make a significant difference and I can't reproduce those times easily here. Would it be possible to get hold of a stripped file that reproduces the issue? It'd really help us to isolate the issue. You can send it to email@example.com
A few things you can check.
- Are collisions enabled? If so turn consider turning them off
- Be aware that Forest Pack processes each pixel on a map so just because one doesn't produce an object, doesn't mean that Forest Pack didn't have to parse it. It can be quite inefficient to create huge dense scatters using very sparse maps. It's much better to use a denser map where you can.
- Is the surface very high poly? If it is, try to avoid enabling auto-update if possible.
- Once a distribution is acceptable, it's possible to Convert to Custom Edit mode which will bake in the distribution and remove any future calculation times.