Feel free to repeat the analysis taking drag into account. You may find it cannot be solved analytically, though. As I recall, it can be modeled as a reduced hole size, but that could be wrong.
h will only diminish when the water outside reaches the top of the cylinder and flows in. Total submersion will follow swiftly, so the time taken for it to reduce to zero can be ignored.
There are three phases:
1. On placing the cylinder on the surface of the water and releasing, it quickly sinks to the point where its weight is balanced by the buoyancy.
2. It sinks slowly, limited by the rate at which water can flow through the hole. In this phase, h is nearly constant.
3. The water outside reaches the top and flows in, quickly reducing the airspace inside to zero.
For the purposes of the time to sink, only the second phase is interesting. The others are too brief.