On crossovers personally I've got mixed feelings on that.
On the plus side it allows more GBs to exist when one model can fit into several GBs. I mean people can only build so many models. I would go so far as to say crossovers help the more obscure GB subjects survive.
On the negative side it seems like it results in less participation within individual GBs because many participants are spread thin over several GBs (I know I have been guilty of this). Instead of one very detailed post, you tend to get multiple short copy and paste posts.
It also results in a lot of duplication as you see the same posts in multiple group builds and occasionally also a WIP under the appropriate topic. This in itself isn't really a problem, but can result in lower quality posts compared to a single detailed WIP type thread.
As to the second, suggestion a topic, but not hosting it. I think there is some expectation that people host topics that they suggest, but a lack of time is certainly understandable. A GB can run anywhere from a few months to a year or more, so much more of a commitment than just building a model for the GB and then going on your way. Ultimately someone has to host a GB, and if the person that agrees to host is willing to run one that was suggested by another, then I see no harm. I've seen many posts over the years where people have just thrown out ideas, I've never seen anyone get bent out of shape over it, and frequently seen people step up and run with an idea another suggested.
I would suggest that those with limited time to host, consider short GBs rather than just abdicate hosting to others. Nothing wrong with a 2 month GB.
Sure it won't bring in the same number of people as a 6 or 12 month GB, but it may see a similar number cross the finish line as those who sign up will probably be more focused. Lots of people sign up for long GBs thinking that of course they will find the time in 12 months, then they actually find they don't.