I don’t have any confidence in Anyone’s reporting unless someone does what that doctor did in Anatolia- test everybody in a community and see how many have the antibodies versus how many show the disease. It could be the 40 to 50% advertised. Or it could be something else altogether. I suspect the reduced death rate compared to previous outbreaks has something to do with the wider testing that seems to be happening this time but general testing is still not done, or none has been reported.
I do share your skepticism at this early point with the reported figures, and your reasoning is sound with regard to the need for more comprehensive testing. I don't think it's possible in that superstitious, backward, third-world environment, but that's a different practical issue.
My understanding is that Ebola typically amplifies and kills more quickly than the body can react and manufacture the necessary antibodies to fight back successfully. A lower kill rate such as is reported for this outbreak could be from any of many causes, but one possibility is that it amplifies more slowly than other Zaire strains, thus giving the victim's immune system more time to react and fight back.
I don't know (and so far have not read) any figures on how many Ebola infectants (virus particles) are required to cause the production of antibodies, and of course that also varies from person to person. But given that it allegedly only requires a very few particles to cause the disease to take hold, it may also be that that same quantity is insufficient to cause the body to react quickly.
So much we still don't know.