I like the idea of "defining the mission"; it does give another direction for thought. It's caused me to wonder already, if anyone has previously considered what sorts of instrumentation would be considered essential for the mission, as I've seen nothing about this here before, and it would effect the final design. One of the first steps in designing any vehicle is to define its payload, if only in terms of gross vehicle weight.
Once you start asking where you want to go and what you want to look at, it seems reasonable to me that you begin to develop different designs for different missions. I think we can all agree that our primary interest would be in systems with a star much like our own, most likely to harbor life as we know it, and these would be the manned missions. Too bad our frail humanity can't stand much acceleration; it makes these the slowest.
We then have choices concerning unmanned missions (depending on budget limits). We could send these to star systems without a star like our own, out of pure curiosity (knowledge is never wasted), or we could send one in advance to a system that we have chosen for a manned mission. This seems good to me, as it would be a pity to send men to an interesting star, only to find that it had no interesting planets. Our astronomy currently isn't up to finding earthlike planets, though we may get there, perhaps with something like a super-Hubble-type device or the 1000 AU "solar lens" telescope. Either way, it will be a slow search.
We might be able to speed up the search, though, with a different kind of remote probe. Pick an area of space with, say, three or four candidate stars, and send a huge-ass telescope halfway there (more or less). The telescope would then have a closer look at all of the candidates and send back information that we could use to narrow our choices.
What think ye?