@gileet84,
This is an area that I have some knowledge so maybe I can handle your "Preassumption" statement. Technically you(or wikipedia) is sort of correct. Some of a radioisotope mineral may leave the system. However, in all reality, this amount isabout 1 in a trillion atoms resulting from abrasion or something else physical.
This is an immeasurable amount compared to the method (We use a specific Mass Spec technique). The "closed system " argument seems to have made its way to wikipedia(I didnt read it yet so maybe its not being Creationist about it).
We dont date "rocks", we date mineral components of the rocks, Minerals like zircon or microlite . These are "calendar type minerals that collect the radionuclides in their LATTICES as they cool" We dont sample sedimentary rocks for radionuclides because the mineral grains are always carried from somewhere else and deposited within that rock.
There are about 45 different radionuclide based dating techniques and [probably the same number of books written on the calibration of specifc techniques for specific rad nuclide tests. Some minerals like Polonium do migrate and give weird results, and these results always show much younger ages than are real. We therefore do intermediate (equilibrium testing of such nuclides to calibrate the measurements)
Only these radionuclides have variable half lives and they dont affect age measurements
1Beryllium 7 can vary its half life by about 1%, its involved in electron capture and is responsible for the only natural occurence of plutonium (Be abuts Uand forms Pu by electron capture) WE DONT USE Be for dating
2Bound state beta decay causes Dysprosium 163 or Rhenium 187 to vary their rates. Thee conditions are so unique as to occur only in meteors and we dont look at using these nuclides for dating weither
3. Weins talks about the "Atomic clock" special case where, if an atomic clock is moving in an orbit the nuclide decay rate will slow down. FORTUNATELY, we dont do any dating using atomic clocks.
HAlf lives and decay rates and decay constants are measured and recalibrated a lot. The fact that the length of years has changed also is immaterial to the method which ses "Annums" which are a compilation of
Quote: For example, the basic equation of radiometric dating requires that neither the parent nuclide nor the daughter product can enter or leave the material after its formation.
Thats not so because we only need to establish equilibrium. By the way, we sample entire crystals that we gather the entire crystal as a sample., whats the harm in an assumption that adds no more than a few hundretdth of a percent error for a method that is available for a fixed range of applications?(For really old rocks and meteorites we use an isochron and concordia tack technique where we compare against fixed and known rates of short decay branches.)
Most mqhinery used today includes Thermal ionization microprobes andSecondary ion Mass Spectrometry where an ion beam is shot accross the polished surface of the mineral crystal and reads the ratios of parent to daughter.
As I said there are over 40 different radioisotopic techniques and they overlap accurately. The errors involved in the method are maximum at about 2% and that translated to a small amount of time (We always report age numbers as+/- based upon the potential error bar)