Quote:

If I put a 20 pound weight on a scale that says it only measures 2 million plus, and I get a measurement of 3.5 million, did I prove that my weight scale is broken?

In your Mt. Helens example all but one sample were estimated BELOW the 2 million year limit. Only one was estimated at 2.8 million and that with a huge error bracket of +/- 0.6 million. If someone really cared about it they could have checked for contamination in either the machine or the rock samples. Given that the ICR people failed to follow simple directions it's not much of stretch to assume that they also failed to remove contaminations (i.e. older minerals) from their samples. More details are here:
http://home.austarnet.com.au/stear/mt_st_helens_dacite_kh.htm

Quote:

A scale is broke, even if its only meant to measure millions, if 20 years can cause it to measure millions.

In Computer Science we call this "undefined behavior". If you ignore the documentation and pass bogus values into a function it might work, it might return a wrong value, or it might format your hard disk. Similar thing with the K-Ar dating: if you put junk in, you'll get junk out.

Quote:

Either way, the earth being billions of years old really doesn't matter to me. Its secondary to my belief that evolution never happened. Whether it didn't happen in thousands of years, or it didn't happen in millions of years doesn't matter to me.

The earth's age may be secondary, but the veracity of dating is not. Because once you accept the later you'd have to deal with the timeline of fossils and why there are simpler ones in old strata and more complex organisms in recent ones.