What would you consider a typical variability in IRR between base and optimistic?

The title says it all -- if you're providing a lot of the assumptions based on a CIM or 10-K and some limited market research, what range of IRR between scenarios would raise the red flags to suggest "this model isn't specific enough"

E.g. right now, I'm looking at a 19% IRR in my base case, but this jumps up to 28% or down to 10% when I switch cases. I've tried to limit a lot of the variables that change in the cases, meaning most change is occuring in the revenue growth rates and margin improvements / COGS. Is this normal or should I try to tighten up my projections and assumptions?