I think one major issue with research as it is, is that it is too easy too just aim for the honeypot straight out the gate, overfocusing on a particular technology chain to get there as soon as possible. Some obfuscation or uncertainty could probably be added to the tech development.