An alternative theory of the statistic might go something like this:
More guns in the hands of private citizens discourage people from committing murder with knives.
The number of murders with knives declined, and the number of murders with guns remained constant.
The overall number of murders decreased because of the decline in knife murder rates, so the percentage of murders committed with guns increased even though the number of gun murders remained constant.
I don't have any idea whether that's an accurate explanation for the higher rate of gun murders than knife murders in the US. The statistical basis for proving or disproving this kind of theory of the effect guns have on murder rates doesn't exactly exist. It certainly is a plausible-sounding hypothesis, though, and no less supported by the lone statistic of 68% of murders in the US in 2006 being committed with guns.
The same guy, in the same comment where he pointed out that more murders are committed with guns than with any other weapon in the US, also linked to UK gun crime figures. Well, sure, let's compare crime rates in the UK with those in the US. We've already established that gun crimes are more numerous in the US than knife crimes, and I'll stipulate for the sake of argument that gun crimes are more numerous in the US than in the UK (though there are niggling holes in that comparison, too). Let's try a different comparison. Note that I'm probably overestimating the UK population and underestimating the US population in these statistical comparisons, which favors the UK in terms of estimating low crime rates (since these rates are measured per capita). The same goes for the fact I'm underestimating UK crime incidences and overestimating US incidences. Despite heavily favoring the UK for determining the per capita statistics, I think you'll find the results illuminating:
In 2006, there were about 60 million (actually closer to 58M, but we'll use the rounded-up number to be kind to hopolophobes) people in the UK as a whole, including Scotland.
In England and Wales alone — discounting Scotland — there were over 163 thousand knife crimes.
By the end of 2006, there were more than 300 million people in the US as a whole.
In the US as a whole, there were fewer than 400 thousand gun crimes.
In the UK, based on these numbers, there was one knife crime commited for every 374 people (rounded down).
In the US, based on these numbers, there was one gun crime committed for every 750 people — less than half a gun crime per 374 people (about 0.4987 gun crimes per 374 people, actually).
That means that, based on these statistics, you are more than twice as likely to be a victim of knife crime in the UK as you are to be a victim of gun crime in the US.
Statistical studies can be great tools for determining the results of policy changes, but the devil lies in the details. Simply picking a number out of thin air — like the fact that 68% of murders are committed by the use of a firearm in the US — in no way proves anything other than that 68% of murders are committed by the use of a firearm. That alone doesn't mean you're in more danger in the US because of laxer gun control legislation than in the UK, where firearms are all but entirely prohibited (hey, at least the police can check them out of the supply room under very extreme circumstances — right?).
Note that even the statistical comparisons I present here are not sufficient to prove a case. There are too many other variables in comparisons between crime rates in the UK and in the US to reasonably expect any real certainty about exactly what effect gun control laws have in either country. A far more reliable statistical comparison for purposes of determining the effect of gun control legislation is, as I pointed out in gun control arguments aren't exactly "rigorous", to compare crime statistics before the passage or repeal of a gun law to those after the passage of the law — say, the three years prior and the three years after. Other factors will come into play, but given enough case studies, trends will definitely be seen to emerge.
If you aren't prepared to produce statistics like that, you aren't prepared to produce statistics that prove anything worthwhile about the efficacy of gun control legislation.
statistical sources:
Bureau of Justice Statistics: Crimes Committed with Firearms
National Statistics Online: UK Population (England, Scotland, and Wales, excluding outlying islands)