RIMM is already at bottom on the other side of the mountain.
I think to be successful, each one has a niche.
Android captures the people that are: cheap, average, and the enthusiasts
Apple captures the people that are: simple, loyal, trendy
Microsoft seems to capture the aesthetic and/or other percentage of the population.
RIMM captures nothing except maybe some die-hard fans or business owners. Maybe for their strong enterprise support (previously), or their trackball/keyboard layout - I'm guessing that represents 0.0001% of the smartphone population.
There are some niches out there that haven't been taken advantage of. One example I can think of are the open architecture folks. Some desktop builders would love an open environment with being able to swap their CPU's and memory and such rather than changing a new phone every year. If RIMM makes something good to capture a good audience and returning their way back up, they have a chance of making a return.