Handling of time series type 2 should be reconsidered

[jeroens]

There is a fundamental difference between how EwE5 and EwE6 handle forcing time series, TS of type 2. These differences have been the subject of discussion in a few cases, most notably in issues #623 and #717.

In EwE5, TS 2 exist in memory and overwrite the data in an actual forcing function when the TS is applied. If an Ecosim scenario with an applied TS 2 is saved, the forcing function is altered forever.

In EwE6, TS 2 are converted to forcing functions during time series import. TS 2 thus do NOT exist in memory in EwE6.

Now, it needs to be assessed if this approach is correct.