Sandiapaul,
Thanks for finding and posting the Greer patent. It does indeed explain the reasoning behind the double head. The purpose of the second head is to provide a new and improved way of pulling the spike, however, it does require a new type of spike puller. I wonder if Greer also invented and patented a puller for his spike.
With a conventional spike, the puller is placed at right angles to the rail with its fulcrum resting on the tie, whereas with the Greer spike, the puller is placed parallel to the rail with its fulcrum resting on the rail base. So the top head of the Greer spike is needed to allow the spike puller to engage the spike in a plane that is parallel to the rail.
The reasons for having the puller placed parallel to the rail are as follows:
1) To make the spike accessible for pulling when it is located in a frog or other instance where a second rail is next to the rail from which the spike is being pulled.
2) To allow the fulcrum of the puller to rest on the rail base rather on the tie, to avoid damaging the tie.
3) To provide a more solid footing for the fulcrum of the puller than would be the case if the tie happens to already be damaged in that spot.
The Greer spike also allows the puller to be repositioned for three different bites on the spike, as opposed to raising the fulcrum of the puller by adding shims during stages of the pull. The three different bites on the Greer spike are under the top head, under the bottom head, and further down on the taper of the shank.
Another feature of the Greer spike is that the taper of the shank next to the rail allows the shank to create a wedging action, which will force the shank against the rail base as it is driven. This reduces the need to make final blows to the head as is done with conventional spikes in order to snug the shank and/or head against the rail base. I believe this final point is the most nuanced as to its pros and cons compared to the pros and cons of the action of a conventional spike as it is driven home.