You really cannot compare two different systems with two different ways of identifying and fingerprinting users against one another and expect them to match.
Are the differences wildly different, or basically fairly close? When you hit peak times do they both increase, and in slow times do they both decrease?
If the trending is similar, that is about the best you can hope for.
If you get weird random spikes from one and not the other, or times when there is a significant drop in one then you may have an issue... but Dynatrace is not a fully functional analytics tool, it's a performance testing and application monitoring tool... it's user identification will likely be basic as it's not designed for detailed analytics analysis...
You also have to remember that various ad blockers and privacy tools will block different items, that alone is going to account for a lot of differences between the tools and what people have installed that are changing what information is received.
