I propose two new terminologies:
Two favorable conditions for the child to achieve very good Chinese (ILR level 3 or above) in both speaking and reading by their teens seem to be for the lead parent/care giver involved to 1) converse pretty much in only Chinese with the child, both ways, and 2) at least one of the parents has good English proficiency, so that the parents are willing to let English slide in the beginning (“open the gap”), thus creating a proficiency differential in favor of Chinese in the beginning, know when to catch up in English (“close the gap”), and has the means to close the gap.
This does not apply for those who pick up more Chinese in adulthood.
If you have more than one child and hope that the younger children also have good Chinese, I recommend much heavier dose of Chinese than English for the eldest child. You need that eldest one to have excellent colloquial skill up through, say, middle school to benefit the younger siblings. (And reading proficiency is needed to complement / reinforce speaking proficiency and to maintain interest.) That was my plan from the beginning. Except, I had no idea whether the whole thing is feasible and how things would turn out, since nobody has done it before AND shared the experiences, as far as I could find. So, my elder daughter’s English did lag behind initially to the benefit of my younger one. But she was able to catch up by the end of 5th grade. Parents really can’t be sure if the first child’s Chinese will be strong enough to converse fluently with the second child all the way through 8-9 years old (for the younger child); therefore, I think it is more prudent to overshoot in Chinese initially for the eldest child, as one can always catch up in English in middle school, as so many youth immigrants had done before.
My younger daughter would have gotten much less practice in Chinese if not for my elder daughter. Afterall, after ~ 8 years old, it is extremely difficult to find fluent Chinese speaking peers to play with, at convenient times and frequent intervals.
What do you think?