Bluenote38,
Since you are trying to capture very fine sanding dust, I suspect that a cyclone collector would work best. One can be purchased for $50 (Oneida Dust Deputy) - $150 dollars (Clear Vue CV06 which includes an inlet ramp to direct air flow into the cyclone).
However, for less money and some time, one can be made. Stumpynubs set out to build his own Shop Vac cyclone, evidently using measurements to maximize separation. It can be found at…
(Part 1) and
(Part 2), although I am not sure how offsetting the cyclone outlet in the shop vac bucket, as Stumpy did, would work. My observation of a cyclone in operation suggests some separation of dust and debris from the air stream occurs within the dust bin of the cyclone. Therefore centering the cyclone opening in the dust bin may be a better approach.
From my research concerning cyclone separators, I recall that the two factors that affect separation efficiency more than any other factors are: 1) the speed that the air stream carrying dust and debris moves (faster is better) and 2) the length of time the dust and debris remains in the separator (the longer the better). Since there is little that can be done to speed up the air flow, increasing the time dust spends in the cyclone is the only factor over which one would have control. The time debris and dust spends in the cyclone can be increased either making the cyclone funnel taller or placing two or more cyclone collectors in series with one another.
These two factors may also affect the separation efficiency of the Thein baffle; that is making the height of the Thein baffle taller may work better that a shorter baffle; but I am not sure.