I think the answers are

(1) B-flow uses coded excitation to improve the signal to noise ratio while CFI uses narrow band pulses;

(2) B-flow uses the weighted decoding filter for clutter signal reducing while CFI uses the conventional high-pass filters;

(3) B-flow uses the envelope of the tissue equalized signal while CFI uses the magnitude of the autocorrelation of the echo.

Hao

you have given a good answer- the best is point 3 in your answer

please see the link hopefully it can help 

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4923974/