Read the article in this month's Inspection Trends on the subject of measuring skewed T-joints.

The Engineer is supposed to indicate the required throat on the contract drawing (structural design drawings). Refer to AWS D1.1-2010 clause 2.3.4 (2). This throat dimension does not include any allowance for Z-loss since the structural designer does not know what companies will be bidding on the job much less than who will actually fabricate the job. Z-loss is dependent on the welding position and the welding process, so there is no way the structural engineer can specify or anticipate the Z-loss. 

The fabricator's detailer is supposed to specify the weld leg on the detailed fabrication drawings. The fabricator determines what welding process is to be used and determines what welding positions are applicable, thu the fabricator must consider Z-loss when detailing the welds. The weld leg must take into consideration the Z-loss per Table 2.2 or 2.9 (tubulars) as well as the weld throat. Refer to AWS D1.1-2010 clause 2.3.5.2 (2)

My position is that the welded skew is not a fillet weld for the simple fact that a fillet weld has to have fusion to the root (D1.1). A skewed T-joint without fusion to the root (Z-loss) isn't a fillet weld. Refer to weld profiles in editions of D1.1 prior to the 2010 editions. That being the case, a fillet weld symbol is insufficient and not appropriate. A detail of the joint cross section and the minimum dimensions for the legs is necessary. Since the leg dimension may not be a metric that can be physically measured, the distance normal to the surface of each member to the opposite toe would be more easily measured by the inspector and welder. See the attached sketches. The detailer can easily calculate the required dimensions and include them in a joint detail.

Best regards - Al