It's certainly to memorize something short term (eg repeat your QQ number over and over for a period of time) but there needs to be spaced repetition over longer periods to make that memorization permanent.
Feel free to provide specific citations, as I did in my parent comment, if you would like to disagree further.
Interesting tactic, feel free to provide something other than your own opinion that it is "irrelevant"
I'd encourage you to actually read the study you're attempting to "cite" by linking to a poorly written Wikipedia summary. Here's the actual link: http://psychclassics.yorku.ca/Miller/
If you read it, you'll find that it makes absolutely no claim about the long-term memorization of numbers outside of the range "7 plus or minus 2" -- instead, it's discussing what it terms "immediate memory", your ability to perform tasks such as mapping specific tones to numbers on the fly, or repeat a series of number you've just been told for the first time, and how your conscious working "scratchpad memory" quickly becomes overwhelmed.
Indeed, the very paper you're "citing" disagrees with you entirely. He notes that subjects can very quickly memorize much larger numbers, but that this does not invalidate his claim re: working memory because this is simply accomplished by chunking the number to be memorized into segments compatible with the conscious mind's small working set.
Again: you have completely misunderstood the paper in question, and are incorrectly conflating "working memory" and "memorization capacity", which are very very different things.
Current theories about long term memory (or "memorization capacity") consider it infinite in capacity.