When you write papers, getting citations and attribution right is part of your job.
The consequences for not doing your job are what they usually are. Depending on the severity and intent, you can get "spoken to" or demoted, up to fired, and banned from "the industry".
Honest mistakes and slips are forgivable, as long as it is not a habit (in which case you're not suitable for the job). Peer reviewers will point out mistakes they see, and you can fix these before they are published. Some mistakes can be fixed after publication with corrections. But if it affects the conclusion of your article and requires a retraction, then it is hugely costly to the journal, and to your reputation. The journal will remember that the next time you submit a paper.
If your paper has problems that are clearly intentional, such as blatant plagiarism, or made-up/falsified data. Then you will probably lose your job. And never be able to get another job in science.
Of course it's not always clear cut where the line is between a mistake and intentional fraud, but a reputation for carelessness isn't good for your career either.
To address the examples:
If a reader cannot find a source you cited, they may contact you. If you then cannot provide the source, people will have good reason to think you made it up. You might be publicly taken to task on twitter, or just become a subject of gossip in academic circles.
Suffice it to say, people will start doubting your other work. They may actively scour your other papers for mistakes, and bring all those up in the open.
I think it's highly unlikely you would be accused of plagiarism for a single sentence that could easily exist by chance. However, surely you didn't make up some person's date and place of birth yourself; there must be a source for it, regardless of whether it's an unattributed quote or not.
If you have a whole paragraph that matches some other paper, or several paragraphs, then it becomes increasingly unlikely you thought of it yourself, and more likely you're copying it. Now, this might still happen by accident, you could even unintentionally reproduce things from memory after having read a paper weeks ago. You may get away with that explanation once or twice, but when it happens habitually people will stop believing it.
Also note that many journals now do automatic plagiarism checks, so there's a good chance your paper (if it has plagiarized content) will simply be rejected, or at least won't get published until you fix it.