Not exactly sure how the exact costs works yet on the mutations, but assume that you have enough mutagen to increase either your base rate by 10x, your clone rate by 10x or both by 5x. Since they stack multiplicatively... note that 5*5 = 25 > 10.
The above assumes that the cost doesn't increase as you increase the multiplier and that they both increase by the same multiplier for a given quantity of mutagen, both of which are probably false, but although these might it might change the optimal point slightly, still leaves you with a setting that except in the edge case where the initial cost is too prohibitive, but since we are arguing in the limit, we can disregard initial costs. For stacking one to be better than splitting between both, the total cost function would have to be strongly sublinear.
Let's play it out hypothetically. Base larva production is 100 a minute and clone is base 100% increase, used clone after production plays out.
Control: After 1 hour you will have 12,000 larva
10x mutation for hatcheries: 120,000
10x mutation for cloning: 66,000
5x each: 180,000
----------
After 5 hours, Control: 60,000
10x mutation for hatchery: 600,000
10x mutation for cloning: 330,000
5x each: 900,000
--------
After 9 hours, Control: 108,000
10x mutation for hatchery: 1,080,000
10x mutation for clone:594000
5x each: 1,620,000
---------------
Now, to compare...
10x hatchery: always 900% more larva from control, no change
10x clone: always 450%, no change
5x each: always a 1400% increase, no change
So, in conclusion, I either ****ed up or none of these mutations have an advantage over the other based on how long you wait. I feel like I have wasted my time.