The issue isn't really even about any desire on the part of companies to abuse personal data. It comes down to the many ways that even companies with the best of intentions can violate this regulation.
It is also trivial to enforce a routine manual scrub of small amounts of private data.
The only people who might have trouble are "small" data resellers having millions of accounts of personal data. It is good they will have to comply.
If you're using a third party service, you will have to use it with prejudice. No more randomly including J. Random noncompliant Ad Agency as a middle man illegitimate handler and washing hangs of it.
Also URLs are perfectly fine as you're not processing the data. The thing that changes is that you have to specify which third party service you use and that it is externally managed. Probably link to their privacy policy and they also have to be compliant with GDPR. This will make it possible to actually get rid of the images and protect your privacy.
Also, why does a company or anyone else "own" personally identifiable information in the first place? The logic is completely inverted and we seem to have all been OK with it in the first place. Why does a company like Facebook "own" what is mine from the beginning — is it just because it is in a row in their database? Is it because I once clicked OK on a TOS? Now I can revoke this consent and Facebook and others have to comply. Previously they could just tell me to FO. Sounds like a great win for users!