How can you determine "falling" of these cases from the link you provided? I checked the progression earlier today and while the situation in Liberia seems to have hit a brick wall, what was not achieved there have been compensated by increased spread in Sierra Leone.
To know if the spread has stopped, you need to compare it to previous data, not just look at the latest.
http://en.wikipedia.org/wiki/Ebola_virus_epidemic_in_West_Africa#TablesThe graphs (above those tables) haven't been updated for quite a while, unfortunately, as it appears Wikipedians are so shallowly interested in it after the media stopped drumming about it. The tables tell everything (and graphs are merely derived from them) but without graphical representation (both in linear and logarithmic) it's difficult to figure the true trends. Just the ever increasing case count doesn't tell much, because it's cummulative to past cases. Only rate of increase matters and since the polling intervals seem to be rather random, you really need to plot them onto a graph to figure out anything from the numbers alone.
Anyway, Liberia is stagnated. Ebola-chan probably burned enough of the population to make it more difficult for it to spread via both population reduction, public awareness and actions taken by medical staff and military.
Still, the last number for Liberia (7,797) is from 9 Dec 2014, not Dec 14 as claimed. It hasn't stopped completely. Correct official number is 7,819 (and this is only confirmed ones, true number may be double or threefold, who knows).
If we would get over this stupid topic of crying about Ebola-chan being lazy, and actually sacrifapped and bled for her to visit India, Thailand or even get loose on Bamako (Mali), we might actually see some development. So stop crying already.