Take a look at what generates how many merits in 30 min and balance it on that.
Except most of those values aren't constants.
Rares depends on whether Ngurii is in a highly specific state as to whether it's "decent" or "extremely good".
Mining is excellent in a few systems and impossible-to-useless in most.
The Odyssey actions don't do a lot in systems without landable planets (but can generally all be done together in those that do)
Scanning is pretty terrible for merits/hour if you don't exploit the bug, but adds to a whole range of Reinforcement actions.
Whether you can get good values for donation missions depends on the station economy and the factions present and their BGS states.
Exobio is far more powerful if you use 3rd-party body databases to quickly get undiscovered Stratum candidate systems.
(etc)
If they balance the actions to their fastest ideal-case value, then most of them end up useless in any other case.
If they balance the actions to their "in an average system" value, then a lot of them will be massively overpowered (like mining is) in the right sort of system, with no guarantee that they'll be offset by a comparably overpowered action for the other side.
They could certainly do better than they have done - and should certainly have planned to need to do more continuous rebalancing than they have - but there's too many variables for it to be possible to balance properly just by tweaking a few baseline numbers.
(For that matter, even having the merit earning rates balanced out to an "ideal" where it doesn't matter what you do, you get 3000 control points an hour for it ... won't solve any of the issues around undermining generally being a really inefficient way to benefit your own Power)