But modern bacteria, as they evolve, become simpler, not more complex. They discard unnecessary DNA, and with it, opportunities to evolve, faster than they add new material, or they tend to go extinct. Complexity is a drain on reproductive speed, and competition is extreme.
By Barbara McClintock's model, they incorporate useful viruses and prokaryotes occasionally, taking a sudden step up in complexity, but they spend most of their time optimizing by pruning, and end up far simpler than the sum of the parts. The combination may ultimately end up simpler than the original cell, if the additional material provides some function of life in fewer letters and all the rest of it can be thrown away.
In a fully-populated-enough environment, this drive to austerity would keep life's complexity cycling around an optimal place, fairly low.
One could claim that this is because they have taken up the niche below the point where multicellularity pays off, and if there were not more complex forms above them, they would change. But actual examples of pre-Cambrian life don't bear this out. Bacteria, as far as we can tell from what little evidence they leave, actually seem to have been similar all along.
There is at least one theory that the Cambrian explosion https://en.wikipedia.org/wiki/Cambrian_explosion#Possible_causes_of_the_.E2.80.9Cexplosion.E2.80.9D required an excess of available resources from some decisive change or outside event, in order to get over that basic trend downward onto a secondary slope upward in complexity.
Theoretically, there are multiple reversal points on the curve of advantage between complexity and efficiency. So, I would say no. And I would bet the model also tops out at some level where coordination reaches an optimum.