I was a technical director responsible, among other things, software development of major systems.
Based on my experience, analysis of trial data is easier and faster to do than development of a new nontrivial application. You apply off the shelf statistical and mathematical code, which you use to implement the analyses called for in the requirements. In this case the requirements are the SAP, which was already defined and could have been used to prepare ahead of data lock.
Yes, there most likely customization to implement the specific relationships from the SAP, and review/change processes to iteratively finalize the final results.
However I can’t imagine a scenario where interim results have not been produced, and that the general TLD outlook hasn’t been discernible from the preliminary data.
I would guess the general assessment would have been available by January or prior, and more likely prior.
So I would factor that likelihood into the possible scenarios of what’s going on with the TLD delay.