Kerbal Space Program

So 45 science node gives us a complete waste of space in the Stayputnik, which really should have a top attachment node, then 90 science gives a useful OKTO with basic SAS and a top attachment node and a better shape.
Continuing on, 160 science in a different branch goes straight to having every kind of SAS available (though you may need a cubic octagonal strut or something to mount it on a rocket), but in the branch continuing from the OKTO, you only get prograde/retrograde hold with the HECS. It's not until the HECS II way down the tree that we get the fly-by-wire's SAS capabilities in a probe body.
Over in the middle at 300 science, you get the OKTO2 with decent SAS capabilities right alongside a heavier probe core in a less-convenient shape that has only basic SAS. Also in the 300 science category, you get the rovemate...which only has basic SAS and is really just a flatter QBE.
Finally, the last part of the probe core tree gives a huge probe core that has an equivalent guidance system to the fly-by-wire and way less torque in reaction wheels than the large reaction wheel module that unlocks alongside the fly-by-wire and an extra .3 tons of mass for just 30 units of EC.

Having a fully-capable guidance system at 160 science, barely-capable ones whose only real benefit is a rectangular prism shape at 300 science, and an insanely inefficient one at 1000 science feels pretty off and makes for some design non-choices; I'll probably slap an OKTO2 on everything until I really need maneuver-tracking (at which point I'll usually slap on a fly-by-wire), and then there's no reason to ever use much else besides HECS II or an occasional Rovemate (which I won't prioritize unlocking anyways).