The term "tricking" can have a lot of different meanings, everyone has their own view or definition. That being said, for the purpose of this post, I'm going to stay the lines of what Matt Cutts always says on panels at conferences, tricking the search engines is "showing the search engines robots one thing, and the user/viewer of the website another."
Even that has some slippery slopes. Most professionals would take that as "don't sniff out the search engine robot i.p. address and show it something different." Doing that can get you banned by the Search Engines. So why all this fuss about CSS?
Any SEO professional who's encountered a website that has been designed by an over zealous graphic artist or web designer who has no clue about how the search engines work will know why there's a bit of fuss going on. However, for those who don't understand what the fuss is about, CSS is a way to help websites that are overwhelmingly designed with images, or use images with imbedded keyword content. By using CSS you can still have the images, but also have some content, with the content being only visible in the actual HTML code. Essentially, this is a form "hidden text".
Technically this is showing the search engines one thing, and the viewer of the website, something else. However, until just recently, the search engines haven't been spidering CSS files. Since this discovery it's raised a lot of questions. The biggest question now, "Is Google getting smarter with CSS files used with 'hidden text'?"
There was some questions about this at Pubcon, and the trio on the Search Pulse (Barry, Chris and Ben) discussed this to a degree a few weeks ago. Ben's advice was to make sure the CSS file was in a folder that was blocked from the search engine spiders. Of course, there are legitimate reasons for using CSS in the manner, the website that is heavily constructed of images that needs to be optimized is the perfect example. However, with anything that has a legitimate purpose, there is always a way to exploit it.
So if the Googlebot is getting smarter and is looking at CSS files to figure out if a site is utilizing the hidden text option, will it penalize a website? What if the CSS file is blocked with the robots.txt file, will the search engine assume something "bad" is going on? These questions and likely a slew more related to this topic are going to be asked of Googlers like Matt Cutts, Vanessa Fox, Amanda Camp and Adam Lasnik, and SEO professionals will be hanging on their every word. If the other search engines send their spiders the same route, I'm sure they are going to see the questions as well.
Eric Enge also has a nice post about this CSS issue to, make sure to read it!