I have an ongoing pair of models of the universe. The first is the one that seems the most probable. The second is the world at it's most plausibly interesting. If there is no direct law of physics against an event, consider it plausible, if only for a second. An example from the latter being a sincere hope that some kind of initiative is conducted to put permanent livable installations on the Moon or Mars or both. I have some theories about the way that such structures could be feasibly constructed, and none of it seems to violate the laws of physics as I understand them. The only thing that I see being a direct problem would be money, and to quote Heinlein

Anything which is physically possible can always be made financially possible; money is a bugaboo of small minds.”

A large reason why I'm excited about building a proper gaming PC, so I can download Kerbal Space Program and a bunch of mods and see if I can create something like the process and architecture that i've been thinking about for a long time. For example, some scientists have conceptualized a band of solar panels that completely encircles the equator of the moon, as a mega source of power. Such a thing is preposterous in the short term, but if the process could be automated, it would become a progressively better investment in money, resources and energy, as each panel that came online began to immediately pay itself off, rather than once the whole array was in place as with traditional Terran methods. In the beginning the power levels would be relatively low, requiring no conceptually new hardware. By the time there was enough energy being generated by the array for overload to be a serious issue technology will have advanced somewhat, and creating/transporting the more sophisticated and robust system to handle the total load will be easier/cheaper.

As a single example.