I think there is a confusion between 'accountability for the school' based on progress from KS2 - ie the calculated Progress 8, which MUST use KS2 data - and the target setting internal to the school, which can use different tests and different measures, but which essentially will be how the school thinks they can best motivate their current pupils and teachers so that they get good Progress 8 results and thus put them safely in a decent position in the DfE data.
So, say, in future 100 at KS2, to get a Progress 8 of zero (ie the national average), should convert to a 5 in the new GCSE [not saying that s what will happen, just for illustration].
The school will know that some children who got 100 won't get 5s - through terrible family circumstances, through over-assessment at KS2, whatever. So to compensate for that, and if possible push their prgress8 above zero, they internally target many children who got 100 to get 6s instead of 5s and set teachers' appraisal targets etc to achieve that.
On results day, they hope that enough children made their 'internal' target of 6 to compensate for those who don't get their 'external' target of 5.
So there are two things going on here - as an overall school, the school will be held to account for all children who got 100 getting 5s. However, in order to maximise their chances of achieving that, they actually set internal targets of more than 5 (and in a few cases might set targets of 4, to recognise the impossiblity of the 5 target).
At least, that's how I understand that it works.