We did build two for MER, but then went back to one for MSL and for the upcoming M2020. The community provided input for this, so it's not an arbitrary decisio

We did build two for MER, but then went back to one for MSL and for the upcoming M2020. The community provided input for this, so it's not an arbitrary decision. The definitive guidance on scientific thinking behind these plans is MEPAG, see: https://mepag.jpl.nasa.gov/reports.cfm https://mepag.jpl.nasa.gov/reports.cfm. Besides its most recent report from 2015, it links to an older National Academies report (the 2013 Decadal Survey). Turning to page 160 of the (long!) Decadal Survey document (free!), we read: "The committee carefully considered the alternative of several rover missions instead of sample return [for Mars]. It is the opinion of the committee that sample return would have significantly higher science return and a much higher science-to-dollar ratio. [...] Existing scientific knowledge of Mars makes it possible to select a site from which to collect an excellent suite of rock and soil samples to address the life and habitability questions..." So, one way to reframe your objection is that we are in truth planning multi-mission coordinated observations, but in the context of sample-return. M2020 is going to select and store samples, and a subsequent mission is planned to return them. That's the kind of progression that seems to galvanize the science community. Aside: Besides the above reports, which (I hope) provide reassurance that all options are being considered, there's also the context of the site-selection process for M2020 - see https://marsnext.jpl.nasa.gov/index.cfm https://marsnext.jpl.nasa.gov/index.cfm. You might enjoy looking at the viewgraphs for the very last talk of the most recent (2017) workshop, which summarize some of the site considerations.