Yes. I do this kind of work for a living, and the issue number two in health (clinical, real world, cohort) data sharing is that it's all over the place: it has to be made sane and harmonized before doing anything with it.

Also the currently popular tools kinda requires as much data as they can for "statistically significant" finds instead of looking for the minimum data required to answer a question.

The number one issue is that data has value to research (due to the funding mechanisms) groups (and hospitals and biobanks) and no one is willing to share their data until they have squeezed every possible funding out of it. Then they're willing to pool it to squeeze out some more, but it can take months or years for the lawyers to agree on what is shared and how and who gets the credit because there is no trust.

And that's just the paper part of it, I won't start on the actual implementation of any data sharing project, which usually also takes months to sort out all issues.