AFAIR the GO-RER in the document was essentially as envisioned prior to SmartTrack. It assumed no fare integration, minimal to no bus integration, no additional 416 stations, ~10 minute peak frequencies (including additional upgrades to the Richmond Hill line), no straightening of Richmond Hill.
It's shouldn't have been a surprise that with that specific configuration showed few passengers transferring from TTC.
A GO Expansion with a stations at most major streets (Steeles, Finch, York Mills, Lawrence, Eglinton, ...), TTC bus connections, full fare integration (free transfers @ roughly $150M/year subsidy; less than the Ford subway plan will require in additional subsidy), and 5 to 7 minute all day frequencies would likely have different results with the same modelling program.