If you self-diagnose with autism, there are four basic possibilities.
One, you self-diagnose and you actually have autism. Result: You get to know more about yourself, maybe solve some problems that weren't getting solved because you didn't know how to define them. Net benefit.
Two, you self-diagnose and, while most professionals wouldn't say you are autistic, you're very close to it and might have been diagnosable as a child. That would put you on the broader autism phenotype end of the spectrum, with autistic brain-wiring but not across the diagnostic threshold. Result: You get to know more about yourself and get to know other people who are similar to you; and while you don't have major problems, identifying as autistic still puts you in a community of people who are used to eccentrics and accept you more than the very difference-sensitive mainstream does. Still a net benefit.
Three, let's say you self-diagnose and you're totally neurotypical--no mental illness, no ADHD, no weirdness at all. Result: You start hanging out with autistic people, and you realize, hey, wait a minute, I don't really fit in here. These people are having experiences I'm not having and I have no clue what they're talking about. From there, you either get frustrated, leave, and hopefully deal with whatever problem wrongly pointed you to autism; or else you decide, well, these autistic people are OK for friends even if I'm not autistic myself, and you make some friends you wouldn't otherwise have had. In many cases, NTs who think they may be autistic have problems with making friends for some other reason; and in those cases, they may even find the friends they want among the autistic people they have got the chance to hang out with. Net result is just some wasted time, not the end of the world, and maybe a bonus to your ability to communicate with the neurodiverse, if you haven't had a lot of opportunity to hang out with the odd ducks of the world until then.
Four--and this is where the problems start--is if you self-diagnose and you don't actually have autism, but you do have some mental illness, neurological difference, or other diagnosable problem that is causing difficulty for you. In that case, the results aren't so benign, because just like an autistic person misdiagnosed as something else, you may try to deal with a problem while presuming a completely different cause for it than actually exists. In that case, you may be leaving a problem unsolved when it could be solved if you knew the reason. Problems commonly confused with autism include social phobia, avoidant or schizoid personality disorder, intermittent explosive disorder, some kinds of epilepsy, depression (especially dysthymia, the long-term low-level variant), attention deficit disorder, or an eccentric personality combined with some kind of mental/neurological difference.
I'd like to address the possibility that people may self-diagnose and use it as a reason not to try to solve their problems. In my opinion, if you self-diagnose and start blaming your problems on your diagnosis, chances are that you would have found something else to blame, eventually, because you weren't looking for a solution when you labeled yourself--you were just looking for a reason to legitimately give up on finding a solution. That sort of avoidance will happen whether or not the person uses some psychological label as the immediate method; he might just as easily have blamed it on bad parenting, being from a minority race, or being average-looking.
This is pretty obvious when it pops up, though, since you'll find someone who isn't saying, "Oh, here's why I do this; how do you guys deal with this, how can I adapt those strategies," as most self-diagnosed people tend to do, but, "I'm autistic; I can't help it. Woe is me." I honestly don't see that second one very much. But I really don't think this problem has much to do with self-diagnosis, since you can use a correct diagnosis for an excuse just as much as an incorrect one; and would be using some other excuse if you hadn't picked autism. The problem here is avoidance and getting the idea in your head that you're helpless to change your circumstances; it's really got nothing to do with self-diagnosis.
So if you're considering self-diagnosis, how do you make sure you don't fall into the fourth category, the misdiagnosed problem that is harder to solve because it has the wrong label?
First of all, just rejecting self-diagnosis altogether doesn't make too much sense; because there is definitely a place for self-diagnosis. Since it's easier to think about diagnosis in terms of physical illness than in terms of neurological difference, here's an example:
We self-diagnose in other circumstances, and nobody seems to think it's odd. For example, last week, I had a stuffy nose, a headache, and a sore throat. I self-diagnosed the symptoms as a cold and treated myself with a decongestant and hot tea. The cold resolved in a few days. We do that all the time, under a certain set of circumstances: The illness isn't serious; it doesn't require a doctor's visit to cure; and we know enough about it to recognize it when we see it. And if you're wrong about having a cold--say, you actually have allergies--then the symptoms aren't so severe that they actually pose a danger to you.
A cold is easy to self-diagnose, but it wouldn't be so advisable with liver cancer. If you had signs of liver cancer, even if you were absolutely sure it was liver cancer and not something else, you'd have to say, "Hmm, this seems serious. I should go see someone who knows more about this than me, who has been trained to treat it." And while autism is not an illness and cannot kill you, diagnosis of mental conditions follows along the same lines of logic as physical ones.
Bottom line with self-diagnosis is, I think, is I'd only recommend it under a narrow set of circumstances, all of which should be true if you're considering actually doing the research and coming to a conclusion on your own:
1. Whatever problems that led you to consider autism are not severe and don't cause major impairment. For example, if you tend to get exhausted at the end of a day of socializing and have to relax, that's not a major problem; but if after twenty minutes you shut down, melt down, and spend the rest of the day rocking, that's a major problem.
2. You're capable of understanding the concept of diagnosis, including differential diagnosis, have a good background in psychology, have enough insight about yourself that you can take a relatively unbiased viewpoint, are able to study a subject on your own until you master it. If you can't do this, there's a great danger you'll mis-diagnose, and that leads to trouble.
3. You want to self-diagnose so that you personally can solve problems in your life that have been popping up because you are different from other people; or you want it so you can understand yourself better. You don't need accommodations at work, at school, or anywhere else; nor do you need outside assistance.
If it's not causing major problems, you don't need outside help, and you've got the knowledge to do it, then sure, self-diagnose. If you're right, you'll gain self-understanding; if you're wrong, then whatever you do have, or if you don't have anything at all, it isn't serious and can be dealt with on your own.
But if you have major problems, or you aren't sure about how to determine whether you should self-diagnose, or you are trying to get assistance with anything, then you need to go to a professional. I know it isn't always possible to do so; anyone too old to still be in school and with poor or no health insurance may have serious difficulty finding a professional; and in some cases, even though there are severe problems, there has been so much mistreatment that seeing a psychologist can only trigger flashbacks and mistrust. In these cases, sometimes there's no choice but to make the best guess and do what you can with the resources you have. It still isn't ideal, though. Ideally, if there are severe problems for which you need outside assistance or accommodation, you should be going to a professional.
Of course, once you go to find your professional and get his opinion, you have to make sure he's someone who is qualified to diagnose autism and to tell it apart from all the other things that can mimic it--and in an adult, not a child--and then actually to treat you with respect and suggest reasonable solutions to the problem of being autistic in a non-autistic world--but that's a whole different subject.