well, one big sort of omnipresent thing is the lack of any kind of real standards in eurorack… i find this makes patching really different than a format where every module is more or less designed by the same person with a single philosophy in mind. you can of course build an all sputnik or an all verbos or an all make noise eurorack system… i actually really like euro systems that are mostly the same maker and have some kind of constant, but then you’re like “oh why am i using this oscillator that doesn’t sound as good as this one that is half the HP”. i kind of can’t stand having a random module in my case that functionally can work with everything else but otherwise feels kind of out of place. when i’m tinkering with my eurorack configuration i am really trying to make something that feels cohesive, which i think helps me patch better.
the other major thing about buchla is that it’s uni-polar… pretty much all 0 to +10V. not having negative voltages makes me think pretty drastically different about patching.
there are “ref” outputs which are different from gates and triggers… “analog” inputs for sequencers make steps determined by voltage value across the entire range vs being advanced per step by gates. the whole arbitrary function generation idea. pulser clocks that have sawtooth cores. really lots of strange inputs and outputs that are unique to the format that make me feel like i’m using a totally different instrument.
and then there’s this… the maths of buchla 
:
