This is something I remember discussing at some point before, and mcwaffles might have already said something along these lines. But since I'm too lazy to confirm, I'll just share my thoughts anyways.
The experiments should have at least some portion of relevance to particular parts. For example, the rough terrain of the Mun from a surface sample leads to the development of the first sets of rover wheels. The thick atmosphere of Eve leads to the development of more in the aerodynamic area, and perhaps some further development of other command pods with better pressure resistance...
Better "pressure resistance?" That's an interesting idea. From what I can recall, the real-world Venera 4 was destroyed by atmospheric pressure on Venus. Is it possible that atmospheric pressure could be an important part of choosing parts? It's interesting to think about. Anyway, assuming that it's real, a 2.5m lander can probably has just barely enough strength against air pressure to survive, which means that it definitely wouldn't survive entry into Jool's deeper atmosphere.
I guess with the tech tree, KSP 1 has a set of parts for certain named categories, even when they don't always make sense. Is a thermometer a science instrument, or a safety tool? Many would think of it as a science instrument, but it's in a section named "survivability," as if it would help you to know "oh, the thermometer is giving a reading of 500 celcius," what would you even do in that situation with a thermometer? Perhaps those thermal guages on your spacecraft could only show up by having a thermometer on your craft? Anyways, I digress.
I suggest splitting the tech tree into two halves: "Engineering research" and "Science research." Considering research in real life, engineering and science are both very important to one another, but neither is actually the other. An engineer can design an engine, but without the science to even provide the theory, where could that lead?
Let's say you want to work up to a Metallic Hydrogen engine. You would probably need:
A very high "thermal" level, to account for the extreme heat generated by such an engine.
A very high "chemistry" level, since the fuel isn't exactly common
A moderate "structural" level, so the engine doesn't get blasted apart by its own force
A high "flow" level, to ensure the propellant leaves the nozzle as efficiently as possible.
Just spitballing, but I'd like to see other people's ideas for this.
edit: didn't quite finish, and accidentally hit "post"