How exactly does neuroplasticity work?

How exactly does neuroplasticity work? We know it works because we have seen from brain scans that heavily used areas of the brain expand to the surrounding areas. Stroke researchers must not believe that it works because they keep coming up with minor tweaks and declare that by exercising in just this way we have proven neuroplasticity once again. This is stupid and lazy of them. How does a neuron signal to its neighbor that it needs help? And why would a neuron drop what it currently knows to take on new functionality? This should be fairly easy to figure out, you can use nanowires to listen in on single neurons or lay a grid across the cortex to listen in. Now if we could just get Paul G. Allen( he created the Allen Institute for Brain Science) to donate prize money. Whomever knows him please contact him about this idea. We should be able to get there in less than seven degrees of separation.
I see two possibilities, do you believe in good or evil neurons?
1. One neuron calls for help to neighboring neurons and they altruistically go to help.
2. One neuron sends out signals that wipe functionality from neighboring neurons and instructs them to copy the sending message.
Not only do I have to ask the question, I have to supply a way to answer the question. If only I ran a research lab.