Why does CPM for different isotopes....

"The higher the number of Electron Volts, the more ionizing the radiation is" --> This is not so clear, at least the way you say it. Here is what I would say:

CPM is counts per minute within a certain medium. If this is a Geiger-Mueller counter, this medium is a noble low-pressure gas such as argon. For a certain decay energy, there is a probability that gamma-ray will interact and ionize the gas which will produce a "count". This probability relates to the type of gas, pressure, and energy of the gamma-ray.

We normally refer to these gamma-ray energies in keV or MeV (thousands or millions of eV, respectively). For scale, visible light is 2-3 eV, UV light is 3eV-1keV, X-rays are 1keV - 100keV, and gamma-rays are normally greater than 100keV. The decay energies we are normally dealing with are 50keV to 3MeV. There are a number of ways these gamma-rays can interact within the gas but the dominant modes are Compton (incoherent) scattering and photo-electric absorption. In general, and this is counter-intuitive, the LOWER the energy of the gamma-ray, the greater the probability of interaction and thus the CPM increases. This is because photoelectric absorption begins to become dominant at lower energies and this mode of interaction has increasing probability with lower energies. This physics of this is there is a kind of "resonance" occurring between the photon (an electromagnetic wave) and the atomic electrons (particle-waves). The binding energies of inner shell electrons are in the 10's of keV and thus as the photon starts to approach these energies (< 200keV), the energy "couples" more efficiently to these electrons making it easier to ionize the atom. In fact, in photoelectric effect, the highest bound shells have the greatest probability for ionization due to this effect as well as these electrons coupling more efficiently to the recoil atom,...,kind of like it is easier to "push off" the rest of the atom as it is ejected.

So, what does this mean for dose? Well, there has to be a conversion done between the number of counts seen in the GM Counter to what would be seen if you had normal tissue there instead. There are some factors that go into this and the calculation is quite complicated. In general, the density of the tissue is ~1000 times higher than gas and the average Z (number of protons and electrons per atom) is about a factor of 2.2 less. The interaction probability will scale linearly with electron density (i.e. differences of density and Z) at higher energies where Compton scattering is dominant (> ~400keV). At lower energies (<200keV), the interaction probability is higher with higher the atomic Z of the target (~Z^3.5) and will remain linear with density. Higher Z of an atom means higher binding energies of inner shell electrons. Back of the envelope calculation: At low energies Z-tissue is a factor of 2.2 less than Z-argon and density is 1000 times higher for tissue than argon gas. So the interaction rate difference should be something like 1000/(2.2^3.5) = 63. Also, at low energies, you can assume that the energy absorbed per interaction is close to the energy of the gamma-ray. For higher energies, the energy absorbed per gamma-ray is usually much less since the interaction is dominated by scattering where only partial absorption occurs at the interaction site and a second photon usually escapes. This is why x-rays can give you a lot of dose because the incident energy per photon is something like 50-100 keV where photo-electric dominates. Higher energies are not normally used in diagnostic x-rays because the scatter and lack of interaction probability reduces contrast of the images and therefore less information to the physician. In general, any x-ray is a dose-information trade-off...higher doses give higher diagnostic information.

OK, so there is a lot here. What to remember:
1) Lower energy is HIGHER CPM. You should be able to see this in the EPA Rad-Net data where lower window numbers are lower energies and therefore higher CPM. (You may also need to scale with window size for this data)
2) CPM in a detector for a given energy needs to be related to the CPM in tissue which should scale linearly with density and depending on the energy of the gamma-ray, will be linear with Z or a functional (Z^3.5) for high and low energies, respectively.
3) Dose is related to not just the CPM (interaction rate) but also to the energy absorbed per interaction. At low energies (<200keV), this can be approximated by just using the photon energy. At higher gamma-ray energies (>400keV) one needs to scale this down by as much as 1/2 of the gamma-ray energy due to scattering becoming the dominant mode of interaction. A combination of both are always occurring so one must always be wary of thumb rules that assume one mode or another unless we are talking about very high or very low energies.

I'm not sure this helps, but maybe this gives people a slight idea of the complexity of these types of calculations and when they read someones post that has not clearly done this type of calculation and interprets data to dose, they may be skeptical of their methods.