The CDC just came up with some interesting stats.
https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html
Graphs don't translate well on here, so in summary, they go through various scenarios based on what they know to date, and then come up with a best estimate (with what seems to be roughly a middle ground to their various scenarios):
Their symptomatic case fatality rate comes in at .004 (.4%) overall, with (as noted previously in the thread, I think) about 35% asymptomatic. Which I believe translates to an overall case fatality rate of .0026 (.26%).
For reference, the best case scenario appears to be .002, with 50% asymptomatic, which puts it at .001 and worst case scenario .010, with 20% asymptomatic, which puts it at .008.
Might want to check my math on all that, but I think it's accurate.