For Passive 1080 sets ou or TaB is definitely best... I have tested with both options and SBS gives a softer overall picture while TaB gives much better detail and sharpness. The math stands up behind this as TaBs 540 lines of vertical resolution is stretched two 1080 lines then every other one is lost resultgin in all original 540 lines of resolution being shown.
As for the arguments about passive and resolution I posted these alreadly:
http://www.avsforum.com/t/1437658/keep-3d-active-or-make-switch-to-passive-3d-tv/60#post_22625913
http://www.avsforum.com/t/1437658/keep-3d-active-or-make-switch-to-passive-3d-tv/60#post_22625968
In short you can argue anyhing you want but passive sets suffer a resolution issue against active not just in numbers but also in actual use where aliasing plays a huge role in the image especially during overlay text.
I personally elieve this happens beuase your brani IS trying to mesh two pitures.... when the lines are complex detail items with easily recognized patterns the brain meshes properly (each image line for one eye between the ones for the other eye) this makes the most sense to the brain as this makes a cohesive pictue whereas if you laid each eyes image ilne on the others you ge ta blurry mismash.
Conversly when you get high contrast (ie text overlays) the image that makes the most sense (on a small scale) is that you are lookign at two sets of stripes and all the black lines and white lines line up on each other. So ratehr than interperating the left eye and right eye being offset vertically, the brain essentially lays the white lines on the other white lines and black lines on the other black lines.... of course this results in the edges not mathching up well and getting nasty edges.
Also the process of combining interleaved images is not like the process of combinging interlaced images on CRTs or even active shutter glasses.... a large part of that process is image retention - the process where by an image sees an image for a while after it's gone. This effectively means that quickly showing alternating lines to the SAME eye is pretty close to showing all the lines to that eye. Showing a constant set of lines to each eye does not allow for image retention to help blend things before they get to the brain.
The argument that there is no vertical parallax confuses me as to why it's important... this is most important when things have a glare or reflection - each eye sees a different glare or reflection even at the exact same height. So when you toss one eyes view of an item out, you are actually msising image data.
Even without glare and reflections per se each eye sees different things on the same veritcal plane.
That said image quality is not based entirely on resolution - anyone rememer the SP4805? Despite being 480p it had an incredible image due to excellent color accuracy and saturation. I watched many hi def videos on it and swear I resolved image detail that isn't technically possible at that resolution. That's largely why many people like the passive image... it's bright and contrasty - which is also the reason TV's demo modes are bright and contrasty... put a bright contrastly 720p TV next to a dim and low contrast 1080p TV and I bet most people will like the 720P set better even at relatively close range.... similar with passive vs active. All depends on what you are looking for.