Maybe you could start by restricting the SK "fix" to reasonable standards:
The MP3 audio format lossy data compression. Audio quality improves with increasing bitrate.
32 kbit/s - generally acceptable only for speech
96 kbit/s - generally used for speech or low-quality streaming
128 or 160 kbit/s ? mid-range bitrate quality
192 kbit/s - a commonly used high-quality bitrate
320 kbit/s - highest level supported by MP3 standard
Seems strange to me that through the years of the developing mp3 specs, there is no utility which can id truncated audio and delete the files, which may be a contributing factor for short duration files with out-of-range bitrates.
How does the fingerprinting process fit into this puzzle? Can't it make any determination regarding audio integrity?
I don't really think this in scope as you've described it, okay you have a very low quality audio track, but if that is the only version you have do you really want SongKongKong to ignore or delete it.
When you have multiple recordings of the same tracks SongKongs Delete Duplicates task will delete the lower quality one and I do plan to provide a few more options for Delete Duplicates such as http://jthink.net:8081/browse/SONGKONG-387 so might fit in there.