DataSource for FAGLFLEXT and BSEG, or New Table in ECC6?

Question

need to create an extractor to have all the information of FAGLFLEXT, because we need to keep the ledger information and the split of the information. However, we need to add 13 fields contained in BSEG.
Therefore we thought to reads the line items table FAGLFLEXA, and then enhace it throught BSEG table.
However, since we are using ECC6 and BI7. It is not support the creation of DataSources for FAGLFLEXA throught FAGLBW03.
Is it an option to incorporate all fields into FAGLFLEXT.
Can we creat a new table group based on FAGLFLEXT, and then adding the coding block extensions to that table -
how does new g/l and the new table group work in parallel? Which is the procedure to do it?
Documentation says we can create a new table group based on FAGLFLEXT --- its the how does it work in conjuction part...for example...the new g/l handles document splitting and one other thing georg referenced last night...will the split documents go into our new table group?
BSEG does not have the document splitter information that we need (it's incomplete data). It's missing profit centers on many items, it's missing the proper split of transactions.
Thanks for your comments.

Answer 1

Here is more information about this post.
Client situation:  Our client is implementing ECC 6 and is using the "New-GL" features.  Because of business requirements, the coding block has been extended (not insignificantly - 18 extra fields at the moment) to accommodate legal, regulatory and management reporting.  The reporting solution includes standard ECC reporting (e.g. report writer, report painter reports) as well as feeds to BW (BI 7).
The Challenge:  Our understanding is that adding all of the coding block extensions to the New-GL tables (ie. FAGLFLEXA and FAGLFLEXT) may lead to performance degradation in the ECC system.  However, we still need to accommodate the requirement to report by the additional dimensions that are not currently included in the New-GL, so our challenge has been to find a solution that minimizes performance issues, while still allowing us to have all the necessary dimensions with which to do the required reporting.
What we would like to know:  How have you handled this in similar situations?
Have you added to the New-GL tables? How many fields? Performance issues encountered?
Have you created additional table group(s) based on the New-GL and then modified that structure to have the new fields?  How does the additional table group work co-incident with New-GL (e.g. does the additional table group receive document splitting information?)?
Have you created custom extractors for BW?  On what basis (we understand that FAGFLEXA cannot be created as a datasource to feed BW)?