<What balance should be struck between private and state run industry?
The more privatized the better. It's still far from perfect as the government is a monopsonist with a near-infinite willingness-to-pay, and there's still plenty of room for pork barrels, backroom deals, and so on, but much less so than with state-run businesses. If you want to see an example of this just look at M-14 trials and the shitshow state-run companies caused.
<Should R&D be handled by the private or public sector?
See above. No gov grants to private research either, as that will just lead to research project grinding to a near-halt and continually demanding more money before they can make any progress.
<If private industry exists, should the state keep them afloat if they encounter financial difficulties?
Nein. If you can't stay alive without gibs you don't deserve to live.
<How heavily centralized should factories be?
Seeing as the industry is privatized, market forces would dictate that. However, manufacturing industries tend to greatly benefit from economies of scale, so we can expect to see factories to be more centralized than not. We don't see this in the current defense industry because they're spread out over as many congressional districts as possible to max out pork, and make them harder to close down.
<Are you concerned about sabotage or air raids, and if so how do you propose we handle them?
That's best handled by the Minister of Internal Security, not me. I'll work closely with him and seek recommendations before making a decisions. This would be another reason in addition to economies of scale to centralize the industry, as you wouldn't have to worry about security in transit nearly as much, and transit in between two locations is where security tends to be weakest.
<How do you retain skilled engineers during the 'off cycle'? (ie. between procurement cycles)
Why is that a necessity? No need to pay money for people that aren't doing anything, let them use their skills in other sectors when they aren't needed. For exceptionally skilled people that are hard to replace, there are options available. For instance, you could pay them a retainer in much the same way lawyers are–a lump sump equal to x billable hours that obligates them to provide you those hours sometime in the future, should you ask for them.
<How do you avoid retards getting emotionally attached to an idea that lacks real merit. (Cavalry is still relevant!!!1!)
As long as I'm in power, publicly shame them for their idiocy and force them to watch poorly animated infographic videos that show why they're wrong. on repeat. But ideally there'd be a system in place that works independent of the guy in charge.
That's a hard one to answer whenever we're talking about a centralized, bureaucratic system such as the military. One possible option is to slightly decentralize the procurement system, e.g. allow individual brigades or batallions to make their own procurement decisions. This allows for the opportunity for different equipment and different doctrines to be tried out, and to see how they work in the field. The problem of course is that this could end up being a logistic nightmare and very expensive to maintain. You'd need to arrange some kind of incentives system in the budget to allow for that–one thought that comes to mind is telling companies to supply their shit in limited quantities to a unit for free, which then tests them, and if the stuff performs well they get considered for the next major contract. Alternatively, units could only deviate from the standard if it's cheaper for them to use their special snowflake stuff.
<How do you ensure cost effectiveness of equipment?
Heavy competition between suppliers. Part of this will mean seeking out contracts from all companies, not just domestic ones. More rapid re-evaluation cycles instead of long, safe contracts that lets suppliers sit on their laurels. Brigades picking out their own shit will help competition as well, as the military will become less of a monopsonist and more like multiple independent consumers.
<How do you avoid mission creep?
That's basically the same as cost effectiveness, so see above.
<How do you strike a balance between primitive but effective technologies with promising but unproven technologies?
Most expedient way is to take the unproven technologies and make them proven. If we're doing the by-brigade procurement option autism highlighted above, then you could also provide incentives for individual units to test out new shit and evaluate it.