In the Quanta Magazine, Jordana Cepelewicz wrote a provoking article
They don't. The purpose of science is to explain and predict the observations, not to provide us with details of pictures that people arbitrarily made up, such as the assumption that the whole life is clearly separated to individuals by some sharp contours.
Well, it obviously isn't and the article itself gives us lots of examples of objects and processes that imply that the boundaries are ambiguous, inaccurate, subjective, or otherwise unreal. The ocean creatures weren't clearly separated. What about humans?
Well, a pregnant woman views the embryo as a part of her body for some time before she gives the birth to the baby. That's clearly not a terribly game-changing moment in the life of the pregnant women (or both the mother and her baby, using the post-birth picture). The birth can take some time: that's why the moment and the boundaries are imprecise. The baby is pretty much the same what it was weeks earlier. We may have laws that define sharp moments and sharp boundaries and make them legally consequential – but they are human or social constructs. There is no reason to assume that Nature finds these moments as important as we do. In particular, it is obvious that abortion (especially late-time abortion) is "intrinsically" an equally harmful act as the murder of a baby. And Nature doesn't have the obligation to make our laws work well. Instead, we should wisely write down laws that agree with the laws of Nature.
Aside from pregnancy, we may consider e.g. hair. Is hair, something where doesn't have nerves inside, a part of your body? What about one of the last hair that is just planning to fall down from your scalp? Now, what about the content of your intestines? All of it surely belongs to the complex envelope of your body, if it is normally defined. But it's composed of the leftover of food, neutral bacteria and yeast, friendly bacteria that may be rather essential for digestion, and many other things. You may arbitrarily go to the toilet and separate the "part of your body" (informally known as "šit") from the rest of your body.
And what about your artificial limbs? Are they a part of your body? They may be as good as the natural ones... or better. Clearly, both answers are defensible.
Now, as I have indicated, we are multicellular organisms that are by definition composed of many cells. These cells cooperate. Our bodies are really societies of cells that were trained to cooperate, reproduce, and divide the labor in complex ways. The "dance" is stored in the DNA itself. While a human is an individual, a human cell is one, too. On the other side, we may also say that humans themselves are building blocks of larger wholes, families, corporations, parties, nations, races, and those behave as unified wholes or "meta-organisms", too. There is some internal exchange of the information. Just like in the case of "humans as bound states of cells", the degree of information that is exchanged in between the parts is much smaller than the total amount of information that the individual building blocks may be conscious of.
I could go on and on and on. There are tons of reasons why the precise "boundaries between the individuals" are imprecise or ambiguous.
The Quanta Magazine also claims that biologists "badly need" to find some boundaries. No, they don't. The idea that there must be such objective boundaries is just an assumption, not a fact, and it is an assumption that seems to be in extremely strong tension with the laws of Nature as we know them. We might say that natural science makes it obvious that the boundaries between individuals are just an approximate emergent concept – pretty much a social if not political construct – that doesn't have to reflect any underlying reality (or at least not an important, objectively valid, and natural underlying reality).
Does Darwin's theory need a definition of individuals? Not at all. Darwin talks about life and genes that are competing and promoting themselves. A gene is really an idea, a pattern, a trait of some systems. Like a meme, a gene may describe just some approximate (or more accurate) skeleton of many cells, their behavior, their mutual interactions, and other things. These are the things that compete. You may really imagine that the competition takes place at the molecular level and the "big bodies" are just tools of the competing DNA molecules in their competition. The subjects in the natural selections are generally rather general ideas or packages of information. There is absolutely no need for them to be identified as "sharply separated objects". And this sharp separation is impossible or ambiguous or a matter of subjective choices. The requirement that it is "sharply separated individuals that compete" is just an additional assumption added to "Darwin's theory" which isn't really useful for the "engine" or "mechanisms" of Darwin's theory to work and that's why it may be (and should be) thrown away.
The Quanta Magazine also discusses "verbs not nouns" (i.e. the definition of individuals in terms of "spacetime not just space" or "processes not objects"). Well, right, at the end, the ultimate things that science describes are properties of processes in the spacetime, not just objects in space. Objects in space are just useful approximations of "processes in the spacetime" in the situations that are nearly stationary or that at least preserve some important characteristics despite the dynamics. But by switching from nouns to verbs, we don't reduce the ambiguities at all. Instead, we might say that the ambiguities get larger.
An analogous problem in physics – which could be claimed to be the "same" problem but this equivalence is just another assumption – is the definition of the observer in quantum mechanics. Again, there is no need for the observer to be delineated by sharp, let alone objectively calculable, boundaries. The argumentation is really very similar to the argumentation above. Separate cells may be observers; humans including or excluding their hair; or their apparatuses etc. may be the observers. A collective of people may be an observer if it is able to exchange the information. What is actually needed for an observer in quantum mechanics isn't a precise sharp shape (in space or spacetime). What is needed is a logically consistent framework in which truth values (and numerical values) are assigned to propositions (and observables). The agent connected with this assignment may be a cell, a human (with or without the content of her intestines or her embryo), an experimental group, or anything else. Such a perspective can be chosen, the laws of physics don't say which choice is "the only correct one", and there's nothing else with this character of the physical law. That's how the laws of physics work. They require the observer to insert some (mostly arbitrary, subjectively chosen) input before the equations may produce the quantitative predictions.
The separation of the world into "individuals" is just a working assumption that may sometimes be useful (and is needed for many laws to work) but that isn't terribly natural or important from Nature's viewpoint and it is silly to claim that such a working assumption is more fundamental, more precise, and more universally applicable than it is. In reality, it is emergent, fuzzy, and has a limited range of validity.