Well you could use the pitch correction in Cubase. Or even just talk-sing the lyrics and then drag them to the melody in the sample editor.
The only real substitute for a singing human voice is another human singing. The computer speech technology we hear today is sample based and sounds much better than synthesized speech (and even then it’s obvious Siri isn’t real). And synthesized singing is much more difficult than speech because of pitch and length variations. Still it is pretty amazing Yamaha can do this quality at all. These are more for special effects or to make purposefully non-human sounding vocals.
I have a couple of the 1st and 2nd generation Vocaloid titles and they were a real pain to use since you had to enter the lyrics phonetically. But their phonetic alphabet did not match what was in any dictionary I had - very difficult to find the right sound at times.
I’m writing a theater piece with lots of vocal parts well outside of what I can sing. But I’ve been mocking them up using Sound Iron’s sampled singers. You only get vowel sounds (or phrases which I don’t need) so you can’t get words. But at least I can hear how the vocal lines sound and fit together. Some times I pretend they are singing in some mysterious unknown language. But their samples do sound good.