Dean said: "I don't think that's correct. As long as the tubing is big enough that resistance isn't much of an issue then putting the filter at the end of the tubing makes for a higher column of water and thus more pressure at the filter, so a higher flow rate. I.e. the higher the dirty water bag is above the filter, the higher the flow rate."
No, sorry, you are misunderstanding hydrostatic pressure, and so was Keith Selbo when he posted his comment in the thread you cited. The filter itself is the rate-limiting stage in the path of the water from dirty reservoir to clean reservoir. You can imagine it like a simple bottleneck. Hydrostatic pressure of a column of liquid can be calculated using the equation:
p = dpg (easy to remember: "pee equals deerogee")
where p is the pressure (Pascals), d is depth from the surface for which pressure will be calculated (meters), rho (the second "p") is the density of the liquid (kg/m^3), and g is the gravitational constant (9.81m/s^2). So, for a three foot long tube (ignoring the water column in the dirty water bag):
p = (0.9144 meters)(1000 kg/m^3)(9.81 m/s^2)
p = 8970 Pascals (Pa) = 8.97 kilopascals (kPa) = 1.3 psi
Because water can, for this purpose, be considered an incompressible liquid, you can imagine the water column in the tube like a rod that pulls down with a force equal to the hydrostatic pressure at the bottom. At the filter, no matter where it is on the tube (the top, the bottom, or somewhere in the middle) the positive pressure from any water above the filter and the negative pressure from any water below the filter will collude to compel water through it at a pressure of 1.3 psi. The position of the filter at the top, middle, or bottom does not matter.
This only breaks down if the length of the tubing below the filter exceeds about 33 feet, if the tubing diameter is extremely narrow (causing a pressure drop due to friction), if air can enter the tubing below the filter (through a leaky fitting, say), or if the tubing is collapsible. If the tubing and all fittings below the filter are airtight, the diameter of the tubing is more than 1/64 of an inch or so, the length of the tubing below the filter is less than 33 feet, and the tubing below the filter is sturdy enough to not collapse, then the flow rate through the filter will be the same whether it is at the top, the bottom, or somewhere in the middle of the tube.
In practice, flow rates might differ due to practical obstacles. It might be easier to have an effective prefilter if the Sawyer filter is at the top, say.