"Data" can be different to mainstream and also difficult to define.
We found that for a large number of pupils using the old style Pivats simply didn't account for the progress we actually saw, as, for many pupils, it's not just about the academic progress.
How do you account for the progress you've made with a child who previously wouldn't go into assembly or the dinner hall, or would only eat certain foods or couldn't play with anyone without it ending in tears, who then can? (Some schools use ieps, others use other established assessments. The ehcp is also a record of progress through objective targets.)
The problem is that some conditions or reasons for having an ehcp are a barrier to learning and so we need to be aware of an measure that progress too. It can be very challenging to show this. Also, sometimes a child makes huge progress with a particular staff group or in a cohort and due to staff changes or class changes, find a transition to the next setting challenging.
I don't think many know the definitive answer to how you measure and account for the data for these situations but it takes a lot of awareness and understanding of wider contexts to interpret data.
WRT autism. SCERTS is very much worth reading about. It's huge and more than an assessment as it also takes into account what the environment and staff are doing too (the transactional support) but is very illuminating when it's done properly. It's huge though and hard for schools to implement across the board. It's consequently and entirely different way of working in class as the academic stuff is then secondary to the emotional regulation and communication skills. The basic idea is that those areas are more of a priority for those children.