I mean negligible in terms of percentage difference.
Hello.
If you found it to be negible then is negible, whether you measure by percentage or millisecs.
What I wonder about, is what factors you think makes it negible. 
I tend to believe that the Objective-C runtime makes it so, or even if the Event Manager can be a bottleneck that evens out any differences.
And now I am curious about the mundane answer! 
I’m with Shane on this one, it was the argumentation about converting that I disagreed. Just an empty command (read:overhead) is more than converting one data into another. It’s building the event, sending it to teh event manager, there it will be parsed again, then send to the target application, reply with an event to the event manager and the actual caller gets it’s an event back as an answer. In this whole process the performance difference in cocoa using an wrapper class or directly in C is indeed negligible in terms of performance. If the C solution would even be 500% faster than Objective-C it will probably influence the total overhead performance maybe by 1%.
Yes, I was using the word “convert” very loosely – I meant the whole process of going from being as AS object to being a C/Objective-C entity.