A while back, when I was playing a hacker in a campaign run by a friend, and he couldn't justify how logic didn't play a bigger part in the matrix skillset (using skill+program rating instead of attribute+skill). His justification was that hackers are inherently smart and a smarter hacker would be better at doing his thing. The best I could offer was that in 2070, regardless of how smart the user is, the program itself is so sophisticated that it takes care of any intelligence issues and the user's skill in manipulating it plays more of a part than the user's intellect. Regardless, big debate later, he wanted to use the optional rule in Unwired about switching all the tests to attribute+skill with the hit limit being the programs rating. I argued for keeping the rules the same saying that if other players wanted to read into the material and felt like learning more hacking, the RAW would at least be reference to everyone in the group who wasn't a rule monger like the GM and I. Long discussion later, we agreed to house rule.
Here it is:
Much like Comm's system rating is limited by its Response rating, and much like a program's rating is limited by system, we agreed that a program's rating is limited by the user's logic. This accounts for a lot of things such as easier to use (and lower rated) programs having simpler interfaces don't allow much altering of parameters to run. They more or less have a "Go" button. Harder to use programs allow lots of tweaking and altering to the essential program through menus or on the highest level, alterations to the code, but unless there is extensive knowledge of how to manipulate these, the user can't get much function out of these bells and whistles. Therefore, a relatively low logic user would be able to run a high program and still hit "Go," but wouldn't know what menus and settings to futz with to get a better job done. This was also a big fix to the script kiddy problem that was a big issue with any matrix discussion he and I had because it meant that any agent you commanded would only work as well as you could logically command it. Therefore, we no longer have the stupid logic 2 rich corp brat buying the best agents to trounce our starting hacker. His agent can only work as well as that logic 2 script kiddy can command (or write script for) him to work.
Aside from the other optional and additional rules we use from the core expansions, this is the only house rule I've ever used, and it seems to keep things well balanced and has played alright for the most part (and adds to the role playing when you have a hacker or rigger who is naturally by his/her stats smart and geeky as opposed to some fluke dumb low attributed hacker relying solely on his programs and hardware).
What do you guys think?