I would agree with your decision, but I see a correct output in my IDE (when kata doesn't see it) + I've written my own kata's test where I reproduced this case. And I see the corrected output. The issue is only a final test...

The Goland sample test is incorrect for parameters then I should show results between both 10,000 and 9,950.
from description:
(Note: m, n, f, b, c are positive integers, where 0 <= m <= n or m >= n >= 0. m and n are inclusive.)

result from wrong test looks like:
Log
Expected
<[][3]string | len:0, cap:0>: nil
to equal
<[][3]string | len:1, cap:1>: [
["M: 9991", "B: 1427", "C: 1110"],
]