A picture of H2 cylinders showed vents. What makes this design poor?

I am not sure which picture you are referring to so I will attempt to answer.

If you are referring to the incident where a fire occurred and the vent system was damaged, then this may have been due to lack of proper supports and incomplete assembly of the test systems. In the past, vent systems were not pressure tested for strength but that is changing. 

If you are talking about small standard H2 cylinders (250 scf or less) then a review of the rupture disc operation is in order. Most individual cylinders have a lead-backed (referred to as fused) rupture disc (CGA type CG4 or CG5) that does not connect to a vent stack. Due to the required mobility of cylinders, a vent stack is not required. The rupture disc is not supposed to vent hydrogen unless the cylinder is engulfed in a fire both melting the lead back and then overpressuring the cylinder/ rupture disc. If engulfed in a fire, the rupture disc opening Is required to maintain cylinder integrity. Although adding hydrogen to a fire is not a perfect solution, it is better than a cylinder explosion and then still adding hydrogen to the fire. 

An interstage regulator relief device not connected to a vent stack. This design is poor and occurs more frequently than we would like. It is typical for a 2-stage regulator to have an interstage safety relief device that is not connected to a vent stack. If the 1st stage of the regulator fails open, the increased pressure can open the interstage relief device, If this is not connected to a vent system (which it normally is not), hydrogen will flow to wherever the relief device is pointed.