After the Paris Convention of 2037, robots in warfare have been globally forbidden in use of primary or secondary combat roles, and across the Earth nations began to disassemble their automated units. The political backlash of reverting to sending citizens into combat has been severe, and scientists, politicians, and military officers have began to look for an alternative; even with advanced warfare, wars are still ultimately won and secured by placement of infantry(*).
(*) This point can be debated for modern warfare, but as it certainly is a popular modern viewpoint in various militaries, we're going to go ahead with it as an assumption for this question.
After a tough political battle, the people of the nations of NATO--by a comfortable majority--have signed off on a once controversial project. With the citizens, politicians, and media (mostly) supporting them, NATO scientists have begun to theorize and develop lab-grown humans to fulfill military roles. They're not citizens, they're not natural, and public opinion accepts this as the 'lesser of the two evils.'
The biological engineering baseline--as allowed by current technology--allows engineers to grow a clone of a human in a chamber. The growing process can be speed up 8x currently (allowing a biologically 16-year-old male to be made in 2 years), however the clones are in stasis during this period, and come out with no knowledge or skills beyond that of a newborn. They emerge with the free will, mental capacity, and variety of personalities of normal humans.
The process isn't cheap currently, at \$1 million (in today's dollars) a pop if left in the vat until 16. Compare this of the figure of costing (very approximately) ~$300k (today's dollars) to raise a child to 18. The scientists reason that this price could be reduced if specific effort went into this effort.
This is a starting point, but things need to improve.
In this Earth remarkable scientific advancements are possible, but all research and engineering must be prioritized; they can't research and develop everything, simultaneously. Further, NATO wants to get these lab-grown soldiers in deployment as soon as possible, with new batches with new advancements deployed as an iterative process. Each generation of soldiers should, theoretically, be better fit to serve than the last.
How should the scientists prioritize this soldier project? What advancements should they focus in the short term, and what advancements in the medium or long term? What would the iterative cycle of new soldiers look like over the decades to come?