A recording system is judged on fidelity. Did the advance post correctly, did the fee calculate correctly, does the ledger tie, can it reproduce March. Those are the right questions to judge it on and the good ones answer them well for decades. Nothing on this page suggests otherwise.
A decision system is judged on something else: whether the person responsible knew in time. It starts from a question rather than a record. Where is the pressure, what moved, what requires me now, what happens if I do nothing until Thursday. To answer that, a platform has to hold not only what is true but what was true yesterday, what it observed, what it could not observe, and what it concluded from each.
The two are built differently underneath. A recording system stores current state and a posting history. A decision system has to store observations as append only facts, versions of every conclusion, the evidence behind each one, and an explicit position on the data it is missing. You cannot bolt the second onto the first from outside, because the history it needs was never captured.
The consequence shows up in headcount and in losses. With a recording system, the work of turning records into a view of the book is done by people, every day, and it is done again from scratch tomorrow. That is why operations of this kind grow staff roughly in line with the book. The other consequence is quieter: the pattern that nobody had time to look for is the one that turns into a loss.