As a linguist myself (working with a few different revitalization projects), you can think about linguistic diversity as being like biodiversity: Examining the differences across many different, unrelated (or nearly so) languages gives better insight into Language (with a capital L) on the whole. Sure, losing an individual language doesn't destroy everything, but each language that's lost is one less (incredibly rich) datapoint which can be used to better understand how people do language, and what other ways things can be done.
For instance, in Wichita, a language which may or may not be dead based on the health of its last few speakers, one could express "the buffalo ran up and down the village several times while scaring people" using a single, very long, very complex word. There are other languages which act like this ("polysynthetic languages"), but Wichita is really, frighteningly good at it. Don't you think that it'd be fascinating to do some MRI studies to see how Wichita people are parsing words, compared to speakers of, say, Mandarin Chinese, which isolate nearly every concept, grammatical or otherwise, into single words?
In addition, as other people have pointed out, when you lose the language, you lose the culture very easily (and vice versa). Even if you're not interested in the specifics of how language works in the mind (or just in general), understanding different cultural approaches to the world provides more information on the human condition. If your culture doesn't permit or believe in the idea of "selling land", that's interesting data, and food for thought for most other cultures.
In short, practically, in terms of trade or war or politics, there's little reason to have a group of 50,000 people speaking three languages rather than one. But if you're interested in how human language, culture, and cognition works, that diversity and those comparisons offer data that a homogenous group would not.