If there is still a remaining significant deficiency with the Vocaloid characters, it is their rather limited mouth movements. Much of this is caused by the character designs typically having a rather small mouth, and a lot of the videos being of somewhat limited resolution. At live events, most of the audience will be a moderate distance from the stage.
Those character designers who have experimented with making the mouth bigger have generally turned their creations ugly as a result: it needs a lot of care to make that work. Also, the range of mouth 'shapes' and the frequency of updating tend to be unhelpful.
In the final analysis, it will be only when a deaf person is able to accurately lip-read a Vocaloid as easily as doing so with a human that the requisite standard will have been reached to be able to say they are then as good as the 'real thing'.
A good step in that direction has been taken by producer AkiGlancy, whose initial trials with lip-synching one of the two Miku parts in this short clip is a lot better than what we have seen previously, and without falling into those traps I mentioned above. The song is an excerpt from Story Rider.
Here, the larger mouth looks fine, is updated fast enough to be quite convincing, and the mouth has enough shapes to enunciate the range of sounds very well. It also helps to be in close-up and available in full High Definition resolution...