Hi Maurizio,
 
I agree that the Olividae are a difficult family; as you point out, the
degree of intraspecific variation is such that it is very difficult to
identify all examples of even the common species. Duclos named dozens of
variants, as did Bruguiere and others, and these variant names can be used
if you have a specimen which matches one of their figures. To my mind,
there is nothing wrong with a label which reads, for example :
 
Oliva sp. cf. Duclos (1835 : pl. 9, fig 2 - var. eximius)
 
(these are not real references, by the way, just a made-up example)
 
If, on the other hand, you feel you simply must put a name to every
specimen you have, then it's a question of amassing as many captioned
figures as you can. Ed Petuch has done a lot in the Olividae, and his work
is full of pictures; I think Dautzenberg did a big Olividae paper once,
too, and there's Ziegler and Porreca (I think this last was Olividae; I'm
not in our library right now). When you have a figure which matches your
shell, your label might read :
 
Oliva whateverana Linnaeus sensu Dautzenberg (1926 : pl. 9 fig. 2)
 
This way, you leave open the possibility that Dautzenberg was wrong in his
identification while assigning a scientific name to the specimen.
Nevertheless, I suspect that he (or she) who seeks to classify the Olividae
ploughs the ocean.