1: \begin{abstract}

2: The interpretation of results of recent $\tau$ decay determinations of 

3: $\vert V_{us}\vert$, which yield values $\sim 3\sigma$ low compared to 

4: 3-family unitarity expectations, is complicated by the slow convergence 

5: of the relevant integrated $D=2$ OPE series. We introduce a class of

6: new sum rules involving both electroproduction and $\tau$

7: decay data designed to deal with this problem by strongly suppressing

8: $D=2$ OPE contributions at the correlator level. 

9: %The impact of complications associated with 

10: %inconsistencies between the existing electroproduction and $\tau$ versions 

11: %of the isovector spectral function is briefly discussed and

12: Experimental complications are briefly discussed and

13: an example of the improved control over theoretical errors presented.

14: %on the determination of $\vert V_{us}\vert$ presented.

15: The uncertainty on the resulting determination, 

16: $\vert V_{us}\vert =0.2202(39)$, is entirely dominated by experimental

17: errors, and should be subject to significant near-term improvement.

18: 

19: \end{abstract}

20: