Usually, in order to impress regulatory authorities, one uses the best limit the method allows routinely.
A limit that meets the governmental requirements is a minimum.
HS analysis should have at least in my experience, a limit of 0.995, unless the analytes of interest are reactive or difficult for some other reason. More importantly, what is the recovery of analytes at levels that are at the bottom 10% of the demonstrated range. For example, if your linearity is over 0-1000 ppm, then what is the linearity of 0-100, or 0-25 ppm?
If the recovery of 50 ppm spike (50 ppm @ 100% recovery is based on the linearity line from much larger levels, such as 200, 500, 1000, 2000, 5000 ppm) is less than 30 ppm then you have a problem if you must accurately report less than 100 ppm.
Follow your regulatory authorities guidelines. (Read their literature!)
best wishes,
Rod