Workflow channels with scenario copies

Hi All, 

In our multi year planning process, most entities need to upload four sets of fixed costs (all loaded by the same team, but sent to that entity's finance team by different functional teams), in addition to all the driver-based costs we're calculating (mostly based on form inputs).  For the four sets of fixed costs, we have four import profiles under the same parent workflow.  The entities import their files and then run a DM sequence from a dashboard to run the validate/load for a range of years they can select.  When they run this sequence, all import profiles (under the same parent) that have been imported are validated and loaded.

We had not set up workflow channels on the different loads, and everything was working as expected.  The different import profiles will use some of the same accounts, but different sets of cost centers.  We did not see any collision of data when the same accounts were loaded, and if a balance that had previously been loaded in one import was removed from the import template and reimported, the data in that account would be cleared from the cube (even if that account was only used by one of the four imports) after the next validate/load.

Now that we have started to copy cube data from one completed scenario (CY) to another blank scenario (CY_v2), so that some changes can be made to CY_v2, we were running into issues with our fixed cost loads in CY_v2 (balances not clearing, other values being overwritten).  This makes sense, and we had not considered loading the fixed costs to copied scenarios.  We would like to be able to only load one of the four imports (if say the G&A load is the only one with changes) in the CY_v2 scenario.

We were hoping that workflow channels on the imports would solve the issue in the copied-to scenarios, but after completing that set up (changing accounts to nodatalock, setting up different data source members for each of the four imports, creating the channels, and applying the channels to the data source members), we are still seeing balances "stuck" in CY_v2 when previously loaded balances to a unique account are removed from the latest upload for one of the imports.  We had also gone back and reloaded the "source" scenario with the channels applied, before copying the data and trying to load only a single template to CY_v2.

Is there a way to use workflow channels so that all data loaded through that channel is cleared and replaced when a new import is loaded to a data source member where the channel is applied, or will accounts no longer in the latest import be left un-cleared?  The documentation on the level 2 and level 3 data units seems to indicate that those accounts will not be cleared, but that does not match the behavior we see in the original loads of multiple sibling import profiles (described and bolded in my second paragraph).  It also doesn't seem to match how a typical, iterative planning process would work. 

Any feedback would be appreciated.

Thanks,

James