"Those who think religion is about "belief" don't understand religion, and don't understand belief." -- Nassim Taleb
Religion has 3 main components: dharma, culture, and belief. Religion gets a bad rap nowadays, mostly because people define it as superstitious beliefs, but this is a very naive definition of a complex phenomenon. It is possible to have a religion without any belief in the supernatural.
Atheism is a political subversion of traditional practices, which are heuristics that benefited the survival of humans at some point in history. Some traditions need to adapt to new environments and stimuli, but it's naive to blindly dismiss all of them. Often, practices are useful before we could fully articulate them or scientifically understand them. For example, quinine was used in the traditional medicine of South American tribes for millenia before we scientifically showed that it can be used to treat malaria. The vast majority of significant innovations come from trial and error out of necessity rather than scientific research. As Taleb puts it, birds don't need to thank ornithologists for showing them how to fly using scientific principles. Everyone knows that's ridiculous, yet when we replace birds with humans and ornithologists with researchers in general, we lose our minds. It turns out that we were really good at surviving for 100,000+ years prior to the discovery of the scientific method and evolution.
However, it's important to not dismiss science. The scientific method is great. It's a formalization of the trial and error empirical methods that we have always used. The problem is in using scientism to create revisionist history that ignores the true geneology of ideas and innovations. Other heuristics for obtaining knowledge of the world existed prior to Academia... in fact, even academics usually don't apply rigorous scientific methods to everything they do in daily life. Everyone uses much old and often biased heuristics to navigate through life practically. Yet biased and nonfactual beliefs are often more beneficial than factual and unbiased beliefs. For example, it's better to tell your children to not trust strangers because they are all kidnappers. If you tell them that technically, less than 1% of people in the world are kidnappers, they would be far less cautious and greatly increase their chance of actually being kidnapped. Even if very few people are actually kidnappers, kidnappers aren't stupid. They play a numbers game. It's not like they would ever approach only one kid, and over time, they get better at identifying which kids are easier to kidnap. Also over time, you child gets exposed to more and more strangers, increasing the probability that he will encounter a kidnapper. Eventually, the kidnapper would identify your naive kid as an easy target, because your "rational fact" has exposed him to greater fat tail risks.
Strangers are all kidnappers is a false belief, but religion isn't about belief. Belief isn't the important part of life that insures your survival and happiness. Atheists and utilitarians are biased towards the obvious first order consequences and dismiss the second and third order consequences that they are ignorant of. Dharmic heuristics often seem stupid at the superficial first order level, but are usually actually important hedges against second and third order fat tail risks. The ancients might not have been able to conceptually articulate why they follow such seemingly absurd traditions to the modern scientific man, but they would still follow the practices to mitigate unknown risks. Sure many practices turn out to be ineffective superstitions, but that's just the process of trial and error. Most solutions you try won't work, but that doesn't mean that you shouldn't try at all. Over a long period of time, the ineffective practices slowly die out and gather a set of antifragile traditions that perform very efficient and practical purposes that often initially confuse researchers.
Dharma is the most practical and useful heuristics that withstood the test of time. Some other ritual practices don't have obvious practical purposes, but they aren't harmful either. These practices form culture. Each culture is a centuries long experiment run under different environmental conditions, so the cultures express in very different ways. For example, Chinese food is very different from French food. However, these experiments are ran in the same universe with the same laws of physics, so there is significant overlap of the dharma of all religions. For example, thou shalt not steal.
We uncover dharma through experience of life and confirm them through the study of history and tradition. The scientific method is a great tool for such examinations, but it's important to distinguish the scientific method (dharma) from scientism (culture).