However, in a sense no experiment is a failure - it's just that some of them yield negative results. Some interesting positives to have come out of this for this formidable example of Homo sapiens before us here is that she is utterly unintimidated by the idea of acquiring a second language and is perfectly willing so to do when she perceives a need and is motivated. Moreover, she is fine with the more opaque vocabulary of the intellectual elite - to her, it is not an anti-language. Finally, like myself, she talks German in her sleep! I have no idea what significance, if any, this has, but find it intriguing.
Another interesting aspect of this is the fact that she really does seem to find the sound of German harsher and more intimidating than the Romance languages she prefers (though Arabic with its plethora of pharyngeal consonants is apparently OK with her). This is from her position as a non-language user, since she was introduced to German, Castilian and English more or less simultaneously. Therefore, that impression seems not to be pure prejudice and it seems probable that German really is a harsher-sounding language. It's not to do with prejudice.
I'm not serious about saying nobody should try bilingualism as an experiment, but i do think it's only likely to succeed if you have access to a linguistic community in both languages. Otherwise, there will be advantages, which our daughter clearly has, but you won't get to easy code-switching between the two languages. That needn't matter, but if your aim is to ensure they are open to two cultures, you have to go the whole hog and involve yourself fully in both, which is not always possible.