This vaguely reminds me of an event last year (March 20 2018? that could be totally wrong) in which GFS insisted on warmer/minimal impact in the 48 hours leading up, contrary to Euro/RGEM, and GFS ended up being correct. And Harvey also back-pedaled bigtime.
The synoptics may be totally different, but I have a Memento-tattoo that GFS can rarely pull a coop.
My 6-12 call for Boston area from Wednesday may burn badly. I'm still skeptical a surface low plows into CT/MA. But GFS/RGEM show it, and others are trending that way.
Why could this be occurring? If you start at hour 0 of tonights 0z GFS, and then scroll back for the equivalent Sat 0z timepoint in previous runs, it looks like the confluence / SSS over Quebec is not pressing down as much in more recent runs... in other words, the prior runs may have over-estimated the confluence?
We'll see, but obviously a terrible 0z suite so far tonight...