i would probably say that mac vs pc is a plausible topic, however, you can't take a fanboy approach, if it were me i'd take a whole hci (human computer interaction) approach
and by that i mean looking at each operating system from all the hci principles (things like how easy it is to use the interface).
they both break a lot of them and you REALLY can't be a fanboy to do it or you sound like someone who can't think and analyse
heck you could even pull other applications into the mix too.
maya vs mac vs pc
but if you're like any other normal person (one of those gamers that are doing compsci because they think they'll get into games not knowing that you actually have to know math and have decent problem solving and communication skills) then you'd find hci stupid and you will refuse to believe any of it or take it seriously because you'd rather not question anything or understand why someone says certain things

doing that though would require a lot of work since it's probably an entirely new concept/way of looking at stuff, but you could easily fill up 6000 words saying they all suck but none is better than the other and explaining why (you can provide graphs and stuff too... if you knew what i know about hci then you'd understand why thats possible)

you could look into inverse kinematics or various other animation techniques as well. and ways to improve it.
equally do the same with ray tracing.
you could do the history of computer graphics, like where blinn shaders came from and the math behind the magic matrices and texture mapping
but that'd be more of stating facts and you wouldn't be able to argue a point.

or you could go down an AI route, and explain various learning algorithms, when they work best, how to make them not work so well.


but yeah... i wouldn't say stwert's suggestions are completely off a plausible essay. you just have to take a realistic approach rather than a fanboy approach.