Remove DifferentialPrivacyParams from MeasurementSpec

[matthewclegg]

Reasons for this proposal:

1. Raimundo, Pasin and Jiayu computed the optimal differential privacy parameters for the different types of queries. There is not any reason to use values other than these. See this doc https://docs.google.com/document/d/1JJA3ufmWgDNblDMKTB2KDLEQjBpgR12kVYVOozjdZ3M/edit
2. If we allow MeasurementConsumers to choose arbitrary values for the differential privacy parameters, then we will not be able to use adaptive composition to compute total privacy budget usage. Instead, we will have to revert to a more conservative approach which would reduce the total number of queries that can be performed.
3. It is envisioned that there may be multiple implementations of Reporting and Planning Front-ends. Can we trust that each of these implementations will use the correct values of epsilon and delta?

Instead of having the RPFE provide the values of epsilon and delta, they should be hard-coded into the Kingdom, and a table lookup for these values should be performed based on the type of query.

[wangyaopw]

We should discuss this in one of the workstreams. maybe in "R&F Planning - Eval Framework & Conceptual Design".