Good question!
Certainly if you could inhibit VP40, you would have an effective treatment for ebola, so the problem is likely that it is difficult to target VP40. Why might this be the case? Most small molecule drugs target enzymes because the active site of the enzyme provides a nice pocket in which the drug can bind (of course, there are plenty of examples of drugs binding to allosteric pockets on proteins, see for example, the non-nucleoside reverse transcriptase inhibitors). Binding a drug into a pocket (as opposed to a surface of the enzyme) makes it easier for the drug to form a number of interactions with the target protein, decreasing the binding constant and increasing the specificity of the interaction. For these reasons, targeting non-enzymatic proteins (e.g. transcription factors or proteins whose primary function is in forming protein-protein interactions) with small molecule drugs has historically been difficult, to the point that they have been considered
undruggable. Because VP40 is not an enzyme, it falls in this "undruggable" class. In contrast, small molecule drugs have been developed to combat the major enzyme encoded by ebola, L.
Of course, small molecule inhibitors are not the only type of drug available. Potential ebola therapies also include antibodies and siRNAs. Antibodies, however, can target only extraceullular proteins, so they would only work against the ebola glycoprotein. siRNAs, on the other hand, could potentially work against any of the viral proteins, yet siRNA therapies are only being developed against VP35 and VP24. This is likely because VP35 and VP24 are produced in much lower amounts than VP40 or nucleocapsid, thus it is easier to inhibit virus reproduction by knocking down these proteins (in other words, it is more likely that the amounts of VP24 and VP35 are limiting for virus production than the amounts of VP40 and nucleocapsid).