Many usages, as implied in Q's post above, are optional & supported as such by good dictionaries - e.g. for 'controversy', both pronunciations rubricated above will be given as options.

For those interested in this aspect of linguistic usage, Longman's Pronunciation Dictionary by J C Wells is recommended - it goes in detail into the percentage [found by survey &c] of language users who use a particular pronunciation: e.g. in the instance under review, it shows that a 1988 British survey showed 58/42 in favour of 'trov' against 'kontr', but this had changed to 60/40 10 years later; adding that 'In US "kontr" the only possibility'.