Currently, the work done in the Paratext Project Interlinearizer cannot be imported into Fieldworks directly, and even then only partially. The Project Interlinearizer can be use to make a highly literal back translation and that output can be exported to back translation project in Paratext and Fieldworks will bring that back translation in as a free translation when the Scripture is imported into Fieldworks. The limitation being that there must be a high level of conformity in the USFMs in both the vernacular and back translation projects, or Fieldworks can not match the back translation to the vernacular correctly.
It is true that Paratext can generate a “FlexText” version of the data from the Project Interlinearizer, but this feature is not able to transfer the parsing or the word glossing from the Paratext Project Interlinearizer to the Intelinear view of FLEx. Parsing or even word glossing between Paratext and Fieldworks are not compatible. As I understand it work on making a conversion process through “FlexText” has been discontinued in favor of integrating Paratext and Fieldworks to the point that both programs use the same databases. So for now, the analyst or translator has to decide if they wish to interlinearize in Paratext or Fieldworks.