I think box height is dependent on a number of factors. Simply using vertical jump height might be accurate in some cases but in those athletes lacking eccentric capabilities it could prove disasterous. You see vertical jump is typically assessed with only a countermovement while box jumps are performed with extreme eccentric loads. As a result an athlete who is capable of big jumps from a countermovement (like those who do most of their training in the weight room) may not necessarily be able to reproduce that performance when dropping from a height. Conversely athletes with amazing elasticity and eccentric stiffness may very well be able to jump higher from a box drop than from a countermovement.
As for heels-first / flat-footed ground contacts and ground contact time I've found that the results are actually counterintuitive. That is, ground contacts are significantly shorter when using a heel first or flat footed ground contact than if they were performed with a toe-first landing. When athletes try to land on their toes or the balls of their feet they simply can't withstand the eccentric loads through the ankle joint so the heel naturally drops anyhow, greatly increasing amortization times.