Great post Sara Anne!
Very well said. It's certainly tough.
In industry, what is sometimes done is to apportion a percentage to each of the categories. For example, from an IT budget perspective, you have three or four categories:
-a. Keep the business running (aka keep the lights on)
-b. Mandatory regulatory/legal/compliance
-c. Discretionary enhancements/improvements/new capability (evolutionary change).
-d. Innovation (revolutionary change)
Many times a and b are lumped together because there is really no choice. Typically, the budget spend is 80% in a, or between a and b. Another 15% get spent in c with the remaining 5% in d. As in your example, d is where the big potential for huge returns is but it is also the highest risk.
The really innovative companies, flip the spend by driving down costs of a and further investing in d (20/80) which then helps them even further drive down a.
Maybe someday a similar paradigm could work with research of this type. Invest is ways to make the ongoing research better/cheaper/faster (not just by cutting funds/people).
Anyway....just random thoughts.
Mike