Peak lift only occurs for a tiny fraction in time. It is that period (dwell) just before and after the peak lift that is important in comparison to cylinder head flow. Therefore (for example), if a cylinder head has a good port shape and flow at .500", why not have a lobe profile x 1.60" that exceeds .500" lift so that the cylinder can take advantage of that .500" point.
Peak lift occurs at ONE point (ex. .500"), but .499" and below is hit twice. Does that make sense on why one may have higher lift than what a cylinder head can flow?
Duration is a sign of how spread the individual timing events are. Do not concentrate on a .050" duration number alone.
Peak lift has NOTHING to do with piston to valve clearance problems. It is the period of overlap.