We could also extend delusions to cover religiosity that is considered a societal/human norm.
Whatever the consequences of that are in practice, on paper I wouldn't take the approach of "It's not a delusion because most people are religious and humans naturally have these delusions", in the same sense that just because 95% of young males are compulsive and obsessed with regards to sex that doesn't mean it's not an addiction.
The example you gave is not valid. Psychiatric illnesses are usually normal behaviors that are magnified to such an extreme that they become abnormal. Religious views are not considered to be psychiatric delusions unless they are taken to an extreme. So going to church on Sunday and believing in god is viewed as a religious belief, but thinking you are god and going to live in the woods as a hermit means that you are probably suffering from a religious delusion.
Anything that 95% of young males do would inherently be a normal human behavior. That doesn't mean such behaviors are good or should be encouraged, but they are not evidence of an illness.
Take your quote -
and then replace "sex" with e.g heroin.
"Most young males think about heroin constantly and crave it, and find it impossible to abstain from using heroin. Most teenage males use heroin and could not abstain if they wanted to."
Add in some significant negative consequences and withdrawal symptoms, and that should qualify as addiction.
I wasn't trying to go through all the criteria for addiction...there are others that I didn't list. There also has to be the possibility of experiencing tolerance and withdrawal. Not every addict may experience all the possible symptoms of addiction, but if there is little evidence of such symptoms in patients then the disorder probably should not be classified within the framework of addiction.
It sounds like the argument for this not extending to sex is that sex is "natural", but just because something is natural doesn't mean it is good for our society and shouldn't be treated as a potentially harmful behavior, and thus worth having diagnosis/treatment for.
The argument for this has nothing to do with sex being "natural". Schizophrenia and depression are natural. The argument is based on the fact that sexual behavior is a
normal part of human behavior and therefore people who engage in common sexual behaviors are not suffering from a psychiatric illness. Some normal behaviors may be harmful, but that doesn't mean they are evidence of a psychiatric disorder.
It may seem like this is an issue of semantics, but it can be extremely problematic to try to pigeon hole certain behavioral disorders into the definition of addiction because they often don't fit, which ends up making it more difficult to understand and manage the illness. So you could classify compulsive overeating as a food addiction, but it turns out that it is actually not an addiction, but rather bulimia. Bulimia has a different cause, a different prognosis, and a different treatment, then addiction. Same with "sex addiction", which may actually be a disorder of
hypersexuality. Recognizing that patients are not "addicted" to sex in the way that heroin addicts are addicted to using heroin, but rather are responding to an overly strong sex drive, completely changes our understanding of the illness and how to treat it.
For the OP and others, how we classify these disorders has important implications for how to treat the illness. It may be reasonable to strive for abstinence in an addict, but it is not reasonable in a patient with bulimia or in a patient with hypersexuality. If we try to convince hypersexual patients that they should be abstinent from all sexual activity, then we are setting them up to fail when they relapse, which is almost a certainty.
There is another dissimilarity between addictions like heroin addiction and gambling, and "sex addiction": while most people agree that the behaviors that heroin addicts and gambling addicts exhibit are problematic, some behaviors exhibited by "sex addicts" may only be aberrant from a religious perspective. That difference has profound implications for treating patients, because the best course of therapy for some patients may be to help them to come to terms with their lifestyle rather than trying to help them modify their behavior.