Why log pathogen reduction for pasteurization validation and not absolute counts? From a product point of view wouldn't some sort of absolute target on the amount of pathogens (count) make sense?
Because pasteurization is a manufacturing processes, not a product, it makes sense to define this process in terms of what it achieves. Namely, reducing pathogens. There may be a number of different ways to accomplish this task, and such a process may be used on a number of different products or batches of product with varying pathogen loads. A definition like this helps by providing an objective benchmark for manufacturers as well as inspectors, and often provide criteria for the specific types of organisms which should be reduced in by these processes.
If I have a feed which is "unclean" (has a high pathogen count) wouldn't it be easier to demonstrate a certain log reduction? Isn't this metric penalizing "clean" feeds in a sense?
Using a validated process during manufacturing does not excuse a producer from performing additional testing on finished products when such quality standards are in place. It essentially adds and additional layer of protection to help improve product safety. Absolute counts are often used in quality standards for finished products, in addition to pasteurization process requirements. For example, quality standards for milk, as laid out in the Pasteurized Milk Ordinance (PMO), define the numbers of total bacteria, coliform bacteria, and somatic cells (per milliliter) which may be present in "Grade A" pasteurized milk. But, these methods don't require the organisms present to be specifically identified. So, theoretically a product could contain a relatively high concentration of a single pathogen, but still meet the quality standard based on absolute counts alone. Since certain pathogens of public health concern can be found in unpasteurized milk, it remains important to use a well defined pasteurization process in addition to meeting absolute count standards.