Well from what you have told until here, the issue happens on EB1 and P4. Probes are Ok and calibration fluid is new.

If you are sure there is no interference's and that you are following right the calibration procedure, then it must be a software issue.

There is no certitudes it comes from the spiff.bin file but it can be a pathway to follow just to make sure that system file is not corrupted.

The 1000mv could come from the fact that this is the max redox range. -1000 - 1000mv which is the value recorded by default for the Touch display graph.