Bacteria change over time both quickly and slowly and both randomly and non-randomly. Teasing this apart is problematic.
Can the measurement of the mutation rate of a single gene be meaningful outside the context of the average mutation rate of the whole genome? Can we make assumptions about the mutation rate of a single gene based on the average whole genome mutation rate?
Can we even measure the whole genome rate even with next gen sequencing technology? This would assume that if you sequenced the same batch of genomic DNA a hundred times you would get pretty much the same answer 100 times. Since I have yet to see any WGS tech provider tooting this particular horn I would assume no.
But even if the sequencing tech was 100% accurate over 2 million bases, is any given sampling of bacteria homogeneous, even from a single well isolated colony? It depends on how the actual whole genome mutation rate compares with the human scales of sampling. It depends on whether bacteria have mechanisms for generating hot spots. It depends on whether bacteria have regulatory networks that can modify the mutation rate. Lastly it depends on known and unknown selection pressures and the conditions you grow your bacteria under. Rich vs limiting media is only scratching the surface.
Researchers like to focus on genetic drift because it seems to be the simplest to measure, but who can say how much selection has been applied to any given gene. Even untranscribed DNA can have many functions, much as punctuation functions in language.
In pathobiology, random mutation may play a role over time, but it is probably too slow to be biologically relevant for bacteria competing for space on hosts, except with regards to tuning regulatory networks for adaptive responses to host condition. Bacteria are much more likely to acquire competitively advantageous genes by theft wherever they are exposed to new genetic material. How much of all the restriction and DNA modification machinery is dedicated to defense and how much to theft ?
What is known is that at least some bacteria are naturally competent. Next Gen WGS sequencing could answer the question of how good thieves bacteria really are.
Interesting topic... When you say "Bacteria are much more likely to acquire competitively advantageous genes by theft wherever they are exposed to new genetic material", are you talking about a new form of horizontal gene transfer or it would be just plasmid or free plasmatic DNA transfer? what would be the mechanism to achieve theft? how could bacteria select advantageous DNA or would it be environmental selection acting after acquiring random DNA?
Remember that genetic drift is not just a function of selection (or lack of it), sampling error in population bottlenecks is also an agent of genetic drift. When someone says the mutation rate of a lineage is high, it may simply be that the ecology of the lineage is such that the substitution rate is high among a lot of internal lineages undergoing frequent bottlenecks. Without getting bogged down in a discussion of the meanings of the terms mutation, substitution, population, or lineage, the point is that the rate of DNA change can be a function of the ecology of the OTU.