If, for example, a sign-language was a system for converting a spoken language into a differently—in this case, specifically visually—coded version of the same language, then it would be a relatively straight-forward enterprise to learn a sign-language. And if this were in fact the case, then there would no doubt be more hearing ASL speakers. And, then, the Original Poster's suggestion that this might be usefully exploited in certain other circumstances would, of course, be true. This skill would be usefully exploited in many other situations.
However, happily, if you believe that variety makes the world a better place, sign-languages are not, in fact, differently-coded versions of other languages. So, for example, neither BSL or ASL are 'signed Englishes'. They are entirely separate languages with their own syntax, morphology and lexis. To illustrate, whereas ASL, like English, has a subject-verb-object (SVO) phrase order, BSL, like Japanese, is OSV.
Because ASL, therefore, is a completely different language from English, it is not a straight-forward enterprise for an Englih speaker to try to learn it. For this reason, there are relatively few hearing speakers of ASL, and nobody learns it for the purposes of communicating in loud situations; it takes years of hard work to become fluent in ASL.