The whole purpose of capturing data is to analyze it. That usually means looking at timing and other measurements. Even if your main focus is the contents of protocol fields, you still often end up looking at physical level timing or at latencies or delays between packets or control signals. For example, you might be optimizing throughput by analyzing the delays between receiving a packet, it generating a full flag, and someone fetching the data.
Debugging or optimizing is very dynamic. As you investigate an issue, your needs change rapidly. Your focus could quickly shift from looking at setup/hold times, to missing pulses/fields, to clock frequencies and duty-cycles. It can be very inconvenient to have to constantly specify what measurement(s) you want at any given moment.
This is where the Instant Measurements and Quick Measurements shine. They answer most of your common measurement questions without any configuration or setups and don't clutter your desktop. They are perfect for those 'Hmmm...that looks odd' type moments. Just point or click-and-drag to investigate.
Other times you are very focused on a number of specific measurements. Dynamic Measurements allow you to select any number of the ~75 available measurements to display. The displayed measurements are dynamically filtered to show only the ones applicable to the signal type under the cursor.