I had been theorizing about the tricks required for achieve a better pronunciation of the Japanese using a Spanish Vocaloid. How I can't confirm them due Bruno and Clara aren't released yet, I'm going to add my supposition in this blog's entry.
Compressed u and w
I think is possible get a compressed pronunciation the [u], [U] and [w] t the aid of the phoneme [B]. Adding this phoneme after the previous should generate a closer sound to the compressed U and W of the Japanese Language.
The Spanish has a clear distinction of the glides, and the co-articulation is almost inexistent in the case of the Diphthong, so in theory the phoneme [j] it shouldn't be enough to induce the palatalization of the consonants. However, the phonemes [j\] and [L] probably can induce it, in my opinion the best candidate should be the phoneme [L] which has a softer pronunciation than the phoneme [j\]. The phoneme [j\] must work better for achieve the pronunciation of the Japanese affricate and affricative consonants as the ɕ [S] or ʥ [dZ].
Softening some phonemes
The Phonemes [f] and [x] probably had to soften with the DYN and raised the BRE to simulate the breathy feeling or the slight aspiration of the phonemes [p\] and [h] respectively.
Also the Spanish alveolar tap [r] can be too strong, requiring be softened. Lowering the DYN or replacing (or combining) with phoneme [l] should work for achieve a closer sound to a Japanese R.
Allophones of the N
The Spanish Phonetic System doesn't distinguish the allophones of the /n/ or alveolar nasal as the Japanese Phonetic System (with exception of the eñe or palatal nasal, the phoneme [J] in both systems). Despite this, those allophones exists, for induce them probably it going to be needed add the corresponding consonant.
For example, for a velar nasal consonant (phoneme [N] in the Japanese system) probably will need add a [g] or other velar consonant as [x] or [k] for induce the alveolar nasal became an velar nasal.