Since the incremental effort is so low, it really doesn't matter much what the original frequency range of a signal is.
It does matter. The waves they measure have so low frequency they've barely measured a single period of it. The higher frequency waves are lost in the noise.
So what should they play? A single 1kHz note is pretty uninteresting.
Compare to the cosmic microwave background - we have no problem converting microwaves to visible light (implicitly doing a frequency shift), but the picture you see doesn't vary despite us having had enough time to collect many cycles of data.
The gravitational background should be similar - even if we collected data for thousands of years we're not going to hear a cool 'bloop' like for black hole mergers because interesting transients just aren't in the signal.