Yeah, that's a good one.  D1.1 is not very clear on that whole thing, and I know the Qualification task group has it on their plate to clarify the whole issue.
If the "groove" side has a gap that is greater than 3/16", that kicks you into a situation in which the root opening exceeds the 5.22.1 maximum.  Some people disagree with this, but if you look at Figure 3.11 (A) you'll note that they show two different root openings (Rn2 and Rn1).  Depending on the specific project requirements there are a number of ways to skin this cat.  It might be easiest to change the joint to CJP in this area, or the gusset could be removed and the free edge built up with weld to eliminate the root gap (or cut a new gusset), or even to increase the fillet weld size for the amount of gap no matter how large it is.  Once again, the project requirements, QA system, etc should guide this.
If it ends up remaining a fillet weld, don't forget to ensure that the fillet weld is large enough on the obtuse ("groove") side.  Though 2.2.5.2 (2) requires that the drawings show the actual required weld size, many detailers/draftsmen/engineers still have the habit of only showing the required weld size for a 90 degree case.  I would first verify how the weld size is shown (i.e. actual required for the actual case or for a 90 degree case).  If it is for a 90 degree case then the proper weld size for both sides of the joint need to be calculated taking into account actual dihedral angle and gap.  You would need to refer to both Table 2.2 and Annex B.

If you send me your e-mail address I'll send you more on this.  kipmank@hotmail.com
Mankenberg