this is a bit of an internet hoax (as far as anyone can tell, this has nothing to do with any research done at cambridge). the interesting thing is that people never stop to think about the limited nature of this lexical phenomenon (i think these things are called memes). it seems to work well in english (although, context plays a large role i think), but what about other languages with similar scripts, like french? what about other scripts? what about hebrew.. don't they leave vowels out in the script? how would memes work then? what about cyrillic? what about agglutinative languages like turkish where words can get crazy long? and obvioulsy this has no relevance to logographic languages like chinese.