And don’t even get me started on why Agile is emphatically and unequivocally not a different thing than Scrum itself.
The most important thing for teams is to establish a rapid feedback loop, and make sure someone is tasked with handling that data. Once you know your customers' needs, you can act to resolve them. If you're using a waterfall approach, you can miss the window, which delays your ability to get changes in front of customers, which delays feedback and increases the chance that you're wasting time.
Before we adopted an Agile-inspired process, we had several projects get scrapped after weeks or months of work because we misunderstood the customer's problem. When we adopted a new process, we built smaller MVPs (took days instead of weeks) and got customer feedback that indicated we needed to make a shift. Many of our features got smaller because we didn't get time to over-engineer them.
The major flaw, IMO, is technical debt. Our product is released on a schedule, but we get feedback from users with alpha products, so we have time to clean up the implementation once we get hands-on feedback from customers.
If we adopted Agile 100%, we likely would rack up more technical debt, and I've actually seen this happen in that same company as management made changes (I became trainer and technical lead instead of project manager).
If you drink too much of the kool aid, you'll get a stomach ache.
As a field we simply don't have actual scientific research on these methodologies (just some joke papers that are as good as toss coin).
I've had much more success with feature driven development (Coad).