In uni myself and a friend used this philosophy to get lab space and desks in the postgrad labs (we were undergrads and normally didn't get our own "private" lab space). We had keycards because we needed access for project work[1], found ourselves some unused space, unused desks etc figuring that if anybody minded we would just apologize and find someplace else. Nobody complained and by the time I graduated we had three desks, in three different labs. Besides the two of us, everybody else in my class had no private labs/desks :)
I did as an undergrad in the US states doing both undergrad and grad course work in Molecular Biology both lab space and computer lab space..keycards, key licenses to apps using apple's keyserver, etc. Even got an unofficial desk in one of the Faculty member's offices.
It was not just knowing those Professors but we did enough 'free' work that we could easily ask for bribes..either pay us or give us space..in my case I wrote some computer macros/programs to use MS Excel to do stats for certain science areas as they had particular methods and standards they wanted to follow in lab data analysis, etc.
Its amazing what you can get away with if you just try.
Of course, like you, we had a reputation for getting stuff done. We got keycard access and permission to store our project work (since it included electronics) in the postgrad labs and we then went to see how far we could take it. We produced some great projects for them in return[1]. We even got a photo (alongside then prime minister) in a local newspaper.
My project partner is currently doing a PHD there and has used the similar tactics to get equipment, lab space and other benefits and his supervisors don't mind because he has opened new research avenues and published papers for them.
[1] the project I linked to was pioneered by myself and my project partner, but has since been used for further projects (and has been professionally rebuilt using expensive hardware - ours cost about €300 in materials to build; the professionally manufactured version cost approx. €10K - but was (is, I guess - I haven't been involved) also much, much more accurate
[1] http://dublindan.posterous.com/things-ive-worked-on-1