Slow day at work, so I decided to plot the data from the Wikipedia article in a different way, because it's quite telling.
So essentially Liberia and Sierra Leone are by far the main contributors to the outbreak compared to Guinea and it's particularly out of control in Liberia, although interestingly the mortality is lower in these 2 countries, especially SL.
I only plotted data from May 27th because that's when the first cases were reported in SL and Liberia.
Great job. I got the same impression following the updates from WHO but it's nice to see it in graphics. I think the mortality numbers are off primarily because the countries seem to use slightly different metrics for how they track and report cases.
The numbers from the WHO are divided in cases and deaths which are each divided further into the subgroups confirmed, probable and suspect cases. Here's the latest WHO report:
http://who.int/csr/don/2014_08_20_ebola/en/
Let's focus on the extremes, Guinea and Liberia. There's one huge difference here: most of Guinea's cases are confirmed with a smaller number of probable and almost no suspect cases. Liberia is almost a polar opposite with large numbers of probable and suspect cases with only a quarter of total cases being confirmed.
Now, confirmed cases are laboratory confirmed so there's little room for error there and the numbers should be comparable. Looking at these numbers, Guinea has a considerably less horrible mortality rate than Liberia: 60% vs a horrifying 87%. So with cases that are known to be Ebola, Liberia's mortality rate does fit with their much more concerning spread of infection.
There's one more thing that sticks out immediately: all of Guinea's cases listed as probable are dead. What this suggests to me is that all these cases were DOA and considered likely to be Ebola but that no tests were done because it would have been a waste of resources. If you subtract those cases from Guinea's total their mortality rate drops back down to 60% again which is almost exactly the same as Liberia's total and close enough to the outbreak average of about 55%.
Apart from that we can mostly speculate. It's possible that DOA's in Guinea but no suspect cases means that Guinea isn't reporting cases that aren't either confirmed or dead but it could also mean that they actually have the resources to test every suspect case and that those numbers really represent all there are. I'm not sure which is more likely. In Liberia, the numbers suggest that most cases are being treated without the resources to confirm and that most of the cases that are confirmed are dead by the time the results are in, but I'm not sure of that either.
My gut feeling tells me that the numbers from Guinea are legit, that their mortality rate for treated cases is the one around 60% and that Liberia is much less able to respond, resulting in less accurate statistics as well as higher mortality and infection rates. This is in line with what I've gathered from reports in the press as well. But I doubt we'll know for sure before the outbreak is dealt with.