In economics, the marginal rate usually tells you about the gain for one more unit of effort.
The marginal rate is usually diminishing the more effort you've already put into something in the real world, but some exceptions exist (especially when a certain quantity of effort is required to get any gain; fixed costs).
The last post was about how exaggerated the attention and budgeting on army helicopters usually is. Even I as a critic do not believe that no attack helicopters is the way to go, and the concept of marginal rates is a useful concept for explaining this reasoning. My writing here about repertoires is another good background for it.
Let's say your army corps brings a squadron of attack helicopters into the fight. Nothing really gold plated, just some average attack helicopters, maybe Mangustas.
The opposing force gets either caught unprepared by those helicopters and the latter have a field day for a change OR the opposing forces are prepared. This preparedness is about vehicle equipment, careful vehicle movement and availability of battlefield air defences (SPAAGs mostly).
The restriction of movements is probably not even the bigger issue here; the expense for battlefield AD (SPAAGs are not effective against modern strike fighters and are almost entirely specialised on countering attack helicopters) is likely the bigger one.
My laziness keeps me from looking up more recent figures, but during the 70's a Leopard 1A4 was priced at DM 1.7 million and a Gepard SPAAG at DM 5.4 million. The 1:3 ratio may have shrunk towards 1:2 in the meantime because of more and more electronics in AFVs, but the basic point is enduring: Battlefield air defences cost a lot, and expenditure for them is expenditure that could have been allocated at more ground combat power if they hadn't to face a threat.
Now back to our imaginary army corps and its squadron of Mangustas; there'll be either some turkey shooting time for them or the opposing force will need to afford dozens if not over a hundred air defence assets to counter this handful of Mangustas (which can appear anywhere in and even beyond the corps sector on short notice).
I think it's likely that some attack helicopters are worth the expense (especially if you already have them) because they provoke inefficiencies in the opposing forces, but it's foolish to expect high returns in shape of tank kills from them. This is not going to happen against a prepared opposing force.
The marginal rate for the first couple attack helicopters is better than their fiscal costs, but break-even is quickly reached and a large attack helicopter force operates at terrible marginal rates of utility; likely often even harming their team more through their fiscal and logistical demand than they help it.
Again; this is the same as with cavalry; high potential - largely if not entirely ruined by countermeasures.