Extant bacteria do not hold the answer to early earth origins of photosynthesis.¹ Or, at least, they probably don’t. Or, maybe, if they do, it is coincidence. Or, maybe they do. Microbial genomics and ecology are limited to answering questions about contemporary and extant bacteria and probably rarely if ever hold the key to understanding anything about evolution on early earth. The unconstrainted nature of bacteria fundamentally limit our ability to answer anything about ancestral bacterial evolution and to use bacteria to answer anything about evolution on ancient earth². There are other reasons we might be skeptical about the ability to use extant organisms to understand evolution on early earth³, but we will focus on questions about adaptations in bacteria.
There is probably little need to review the biology of photosynthesis in order to discuss information theoretic limits in extrapolating back billions of years. Nevertheless, just for completeness, here is a figure stolen from wikipedia…

Without reviewing too much basic biology, there are two photosystems involved in oxygenic photosynthesis on earth today (though the article covers a slight modification that they lovingly call photosystem 1.5). From the article,

But in terms of function and structure, the photosystem reaction centers fall into two categories that differ in almost every way. Photosystem I serves mainly to produce the energy carrier NADPH, whereas photosystem II makes ATP and splits water molecules. Their reaction centers use different light-absorbing pigments and soak up different portions of the spectrum. Electrons flow through their reaction centers differently. And the protein sequences for the reaction centers don’t seem to bear any relation to each other.

The article in question discusses Heliobacterium modesticaldum which apparently is the “simplest known photosynthetic bacterium”.⁴ The relevance to evolution on early earth can be summarized best by the researchers quotes.

…H. modesticaldum has presented researchers with an interesting piece of the photosynthesis puzzle. The only photosynthetic bacterium in a family with hundreds of species and genera, heliobacteria’s photosynthetic equipment is very simple — something that became even more apparent when it was sequenced in 2008. “Its genetics are very streamlined,” said Tanai Cardona, a biochemist at Imperial College London.

Taken together, “heliobacteria have a simplicity in their organization that’s surprising compared to the very sophisticated systems you have in plants and other organisms,” said Robert Blankenship, a leading figure in photosynthesis research at Washington University in St. Louis. “It harkens back to an earlier evolutionary time.”

Its symmetry and other features “represent something quite stripped down,” Redding added, “something we think is closer to what that ancestral reaction center would have looked like three billion years ago.”

These quotes assume that “simple” suggests “ancestral” ⁵, but also, “simple” can be “stripped down” implying that the ancestor was complex and this is derived to be simple⁶. The article also mentions that this is the “only photosynthetic bacterium in a family with hundreds of species and genera” implying that it has derived photosynthesis (probably from HGT). That it is the only photosynthetic bacterium in the family is probably not accurate, but there are unique aspects to its photosynthesis that have made it the focus of many studies ⁷. Similar statements about simplicity and ancestry are also found in the original work ⁸.
Despite these complaints, let’s stop picking on this article and, instead, discuss why researchers probably should not use extant bacteria lineages as models for evolution on early earth. Whether in aerobic or anaerobic environments, in deep sea vents or sea beds, whether extremophiles or possessing other strange life histories, it is probably dubious to use extant bacterial lineages to provide insight into early evolution on earth ⁹. There are several reasons to be skeptical, one being evolutionary and ecological in nature and the other being information theoretic. The evolutionary and ecological reason that these connections are difficult is simply that early earth dynamics (e.g., climate, atmosphere, etc.) were different than the cretaceous, eocene, quaternery, and today. The selection pressures and constraints billions of years ago were very different than today. It is therefore conceivable that if a bacteria gained some innovation (e.g., photosynthesis) after early earth, the conditions under which it were gained were sufficiently different as to suggest that the adaptation/innovation would be different in nature. The alternative to assuming that the evolution of a new structure would be similar to a structure billions of years ago is that the adaptation / innovation may be a relic from early earth. This brings us to the other problem.
Information theory is a branch of science that deals with transmission, storage, and quantification of information. While used in everything from communication (of course) to cryptography, it also is helpful for understanding the limits of certain systems in transmitting information. For example, without getting into the math, one can intuit that a two state system (0’s and 1’s) would have less potential information than a four state system ¹⁰. We might also assume that, all things being equal, the two state system will be informative over a short period of time as we might start in a state 0, transition to another 1, and transition again 0, effectively erasing the first transition from 0->1. This is called saturation and will happen faster in a two state system than a four state system. Of course, information theory is significantly more rich than this but this intuitive understanding of information helps conceptualize potential limits over deep time. In general, this is why researchers may use larger state spaces for deeper evolutionary inferences (e.g., amino acids instead of nucleotides).
Constraints offer another dimension complicating evolution at deep times. Constraints present barriers to saturation in that they may slow evolution. Selection acts on variation and a constraint limits variation. Strong selection on constrained structures can therefore result in the evolution of elaborate structures that compensate for the constrained structure. Constrained structures and systems, therefore, may be better for looking at deep time precisely because they prevent saturation, may require elaboration, and may require the addition of new structures onto existing constrained structures. As a result, an extremely constrained population or lineage is more likely to look like its ancestor. An unconstrained and saturated population may or may not look like its ancestor but there is no way to tell as nothing remains from the ancestral population. Another way to express this would be that it may be easier to date aspects of a cathedral than a simple wooden shack. Nevertheless, as mentioned above, even if constrained, the environment and adaptive pressures are unlikely to be the same today as they were 3.5 billion years ago. And so, the constrained structure may still have functioned differently in the past.
Now back to intersection of bacteria, the beginning of photosynthetic life on earth, and information theory. These researchers are examining a complex adaptation/innovation (photosynthesis) but still one that has changed significantly over the last few billion years. As stated by the original article:

Despite the low sequence homology of core RC polypeptides from different organisms, reflecting billions of years of evolutionary divergence, all of these proteins share structural similarities within their central ET domain

However, the HbRC has had a long time to evolve from the ancestral RC and has certainly acquired unique features that are advantageous to the organism.

There is no evidence that these bacteria are constrained, making saturation a hard problem to overcome. They may lose, regain, and lose again the ability to photosynthesize. Also, they can, and probably have, gained photosynthesis through horizontal gene transfer complicating things more. A fast life history (i.e., a bacterial lineage may have several generations over a short period of time) amplifies the potential for saturation to be a problem as the rate of evolution will generally be high when compared to an organism with slower life history (e.g., a human or a tree or an elephant) even if mutation rates and other population dynamics remain the same. Finally, again, even if the photosynthetic pathway is similar to the one that evolved a billion years ago, it may still function very differently despite saturation as bacteria are complex integrated systems and are interacting today in very different ecosystems¹¹.
Nevertheless, the authors of the Science article conclude with:

However, some traits of the last common ancestor of all RCs may be preserved in the HbRC as a result of its host’s anoxic niche, which has similarities with early Earth.

That they conclude the article with this bold statement as well as the above discussion suggests several questions. Are microbial researchers aware of the limitations of their research but push the stories anyway? Or, are they completely unaware of the fundamental limits of their study system? Cynical or not, it is unlikely that this system is a window into the world of early earth. Instead, this is a window into the world of this particular bacteria’s interesting photosynthetic machinery. And perhaps that is still a worthwhile endeavor.

TLDR, the recent adaptations of bacteria may not be the place to look for clues to evolution on early earth.