This is an interesting analogy, thanks.

But, doesn't this match the parent comment's point? If someone asks why a particular point and not, a nearby point is on the fractal, the reply is just that complicated program generated this number and not another number. It leads to the theme of incompressibility (is there a program of smaller length which can correctly check whether a point in a given region is on the fractal). Sometimes, programs are incompressible, and at other times there is a local explanation.

So, people writing on this topic should be careful to make this distinction and check if their local explanation (which is a 'compression' of the evolution process) actually classifies points into fit/unfit regions with the assumption that the points are in some common region(shared features in the environment).

Compare also with explanations of historical events(list of causes for a victory but with the same causes present, there is failure in another context) and atheist - religious debates on God's grace for a particular prayer and allowing the suffering of others. If a local explanation can't distinguish events, then the global explanation has to be validated by something else.

This analogy doesn’t really answer it, it only shows that there’s no generic answer. So “fittest” basically means “the result”. It’s a definition, not explanation. But it is used instead of “the result” because we have some insights into how it was calculated, and we know that the nature simply creates {mi,bi,tri}llions of variants and then some minuscle amount has better chances to pass and mix these genes, outplaying others in this game. That’s where fittest comes from. Any attempt to explain it fully meets with a local complexity, in which only high-level questions can be answered in the same limited high-level manner. So there is some compressibility, but it’s clearly jpeg. Cause if a whole planet of distinct particles would strictly follow some rule to compress the observations into, it would be a rule of physics, not biology. E.g. why do animals fall to the ground? Due to gravity, here’s <formula>, works for all particles.

Sure, you won't get exact predictions on which genetic codes are realized. Similar to how statistical physics doesn't predict microstates, but the macro picture is understood. Even that might be too ambitious for complicated systems, but you can still have sensible predictions from evolution theory which looks at a landscape and says that there will be animals with tall necks and predators with very fast speeds even if you didn't know beforehand about giraffes and cheetahs.

In the fractal analogy, the compressed theory does not predict that exactly this point does/does-not lie on the fractal, but predicting whether a small disc around the point intersects the fractal.