Looking at two different techniques by the following artists - the former is generally higher polycount.
"Falling Pixels" models (Looking at the Fiat 500C); between body panels use single iteration and sometimes 2 iteration chamfers (for very sharp edges) on the base mesh. A Turbosmooth is then added. The benefit of such detailed chamfers is that seperate smoothing groups on the base mesh is not necessary. The detail looks great but with 2 iteration chamfers coupled with 2 iterations or more on the TS, to me the poly count seems a bit high.
"AXYZ Velocity Cars" between body panels and where panels are welded, on the base mesh use hardly any chamfers, though are seperated by smoothing groups. The TS (with seperate smoothing groups ticked) generates a nice result on the bodywork, though when rendered, the detail between panels isn't as nice. (With non chamfered edges).
I basically want to know what the experts think is the current preferred method. I suppose if computer power isn't an option, it's the first one, but I suppose (selfishly) I just want to know what will sell better!!
Thanks for reading, I value your responses!