Multiples (1..n) of the sequence length should be used for the length/duration field to make sure that the repeated sequence (when played back by other software) still is periodic. It's already there for perodic noises but missing for periodic sines (sines with "lock frequency to RTA FFT" checked) and multitones. And more than one digit after the decimal point would be nice, to display meaningful times for short sequences at high sample-rates.