I've been back through the cycle 7 numbers (our last preparation, and the first cycle I have detailed numbers for) carefully. These are my conclusions; it is entirely possible I'm wrong, not least as I'm drawing on only two sets of concrete data, and there's definitely been some funny business with overhead before the latest formula.
So. I'm now increasingly sure we do pay preparation costs per system in this cycle; they're deducted from current income, while the list of what we're allowed to 'spend' is based on last cycle's surplus. On the plus side, I think we only pay overhead & upkeep for current control systems & expansions, not preparations, so it actually evens out mostly.
I can't see any obvious relationship between income and preparation cost. Since we can't see what exploited systems will bring in until they're exploited by a control system as far as I can tell (which may be related to preparation cost), I'm out of ideas on that one. Will just take the numbers as given, for now.
We do get income from systems that enter turmoil this cycle, but not next cycle.
So lifetime of a new control system;
cycle 1 - final surplus sets how much we can prepare in cycle 2
cycle 2 - (preparation) - based upon previous cycle surplus, we have a list of allowed preps by cost. If successful (?), we pay the preparation cost out of cycle 2 income. If paying this cost would take us into deficit, that prep fails. Possibly order is highest ranked preparation goes through first, until our entire valid list is paid for, or we'd go into to deficit to pay it.
cycle 3 - (expansion) - Those that were successfully prepped can be expanded. Successful expansions require normal upkeep & overhead, but we don't get any income from them. If upkeep+overhead would put us in deficit, expansion fails. Which fails first, expansion or preps? Noboody knows!
cycle 4 - (income) - we pay overhead for every control system. We get income from each. We pay 0 upkeep if fortified, normal upkeep if unfortified or cancelled, undermined upkeep if undermined. If we can't pay all upkeep from income-overhead, it puts systems into turmoil; we still count the upkeep for balance next cycle (i.e. we'll stay on negative balance), but it counts up systems, highest upkeep first, enough to 'pay' the deficit, those systems enter turmoil next turn. Undermined most likely to be first.
cycle 5 - (turmoil) - Systems in turmoil cost no upkeep, nor give us any income; we still pay overhead, though possibly only if we can afford it. If still in deficit at end of cycle, systems are lost; otherwise they return to 'normal' next cycle.
cycle 6 - bailout!
Simples.
