I have done a few projects where the editor captured from the Sony Z1 HDV camera into FCP as downconverted DV. Then the project file was reconformed on my FCP from a Sony Z1 camera as HDV, then upresed to Uncompressed HD. All was in sync so yes, the TC is accurate. I have also tested the Miranda and Convergent bridge boxes and recaptured frame accurate from HDV straight to uncompressed HD.
I also understand why you want to use this workflow as editing DV is bread and butter for most FCP systems, even on laptops/G4’s and older FCP versions than 5. It also simplifies the process of making vhs dubs or DVD’s for client approval, without downconverting from HDV. And I thought HDV was the same data rate as DV so same disk space anyway.