Possibly flag-able article, preferring to comment... Given that there's no way of telling what a young PhD or post-doc is capable of performing, why not try to diversify the field. Broadening the pool from where ideas might come and increasing the pool of PhD students increases the likelihood of success in the next generation.
If you truly want to go by "their work", then the only people being funded will be those probably too old to actually do the work. Which merely leads to sending money to the PhD's and post-docs mentioned above with no obvious track record. Check how much work really comes from the senior professors, Nobel laureates, etc. They, unfortunately, spend their time running departments, soliciting funding, giving lectures and serving on scientific panels.
