The current Bosnian locale (bs) does not have enough data in it to make it into ICU. However, although bs is essentially the same as sr_Latn, in many contexts it is politically unacceptable to have region "Bosnia" listed under "Serbian" language: "Serbian (Bosnia)".
Can we take a shortcut toward fleshing out the bs locale by initially populating missing data with items from sr_Latn? Nearly all of the existing translations in bs are the same as those in sr_Latn, but differ in capitalization conventions used in CLDR. The main difference between the current bs data and sr_Latn is in (1) date/time formats (unconfirmed in bs) - and (2) number symbols. It should be relatively easy to check and confirm whether these differences are really intended.