Time management is something that gets built best by doing mocks. For this, you can even try unofficial mocks.
1. The thing with DI is that the trick to doing a question in the most time-efficient way varies from question to question. For instance, graph questions often ask you median values where the most time-efficient way to calculate it might be by making some approximations, but if you start calculating the actual median using values in such questions, you’ll lose a ton of time. So you can build that wisdom by practising more questions and training your mind to think in the direction of "How do I solve this in the most time-efficient manner?" With practice, this thought process should come naturally.
2. As far as MSR goes, the only tip that I can give that has helped me improve is reading it like RC and not CR. You shouldn’t focus on the details of it, rather understand what it is about, where what information is, etc and get a structure of it, so you can quickly come back and read it more thoroughly as a question demands.
3. Sometimes, based on personal experience, I’ve seen you can spend way more time on an MSR/Tables question and still overall time well on the DI section by cutting time elsewhere. For example, you can solve DS questions more quickly and thus save time for other question types.
4. Knowing which question you should skip by guessing is also important. Again, I feel mocks really help with this dynamic decision-making ability.