Sure, such a company would have no chance of succeeding in the stock market, but what if it never plans on IPOing in the first place?
Buffet's core complaint is about companies not making enough profits due to their profligate spending on other things. A profit cap would not have the impact you're hoping for here. :)
> such that all net revenue in excess of some fixed amount is forced by the corporate charter to be turned into either employee compensation or stockholder dividends.
By definition, only profits can be paid out as dividends, so again, a profit cap would prevent the thing you're you're trying to boost.
Er, sorry, I rather meant that profits after dividends would be capped. You can do anything you like with the money—other than keep it in the corporate coffers (or in commercial paper or anything else that's still liquid assets on the balance sheet.)
But yes, you're still right, it wouldn't have the correct effect.
How about a sliding window cap on non-compensation spending, together with a sliding window cap on headcount? So, in a year with record 10x net revenue, you would be allowed to 10x your salaries/bonuses/dividends, but you wouldn't be allowed to multiply your capital costs, or "new" labor costs, by more than, say, 1.3x. (And keep the fixed profit-after-dividend cap, because otherwise the corporation would just hold all its money in the bank until the sliding window grew enough to let it spend it.)
Also, presumably you're talking about new laws in the US here? What stops businesses from just moving their HQ overseas to places with different laws they like better?
That seems likely to cause the exact _opposite_ effect of what you intend. Apple is sitting on a pretty big cash pile right now. If the only legal options they had were to invest it or return it as dividends, the original article's premise is that they would find ways to invest it (buying factories in China, bringing app development in-house, etc) rather than risk the shrinking of the bureaucracy.
The problem seems doubly bad with Wikipedia and other non-profits. The problem isn't that they have too much profit. It's that they are growing their expenses to match income (rather than capping expenses at some "rational" level and trying to maintain enough income above it). In theory, in for-profit companies shareholders will start to complain if expenses increase too high -- it's not clear who would make the same complaint at Wikipedia (other than the original article).
Trying to come up with piecemeal regulatory "solutions" like this is like trying to invent a perpetual motion machine by coming up with increasingly more complicated contraptions. There are fundamental, local interactions that you have to contend with that basically rule out accomplishing what you want.
Everything from the growth of the welfare state, to populist revolutions to put socialists into power, can be blamed on how common this misconception is.
The problem with applying top-down planning to economy is not that it doesn't work; it's that it optimizes for the goals of the few doing the planning, leaving large swaths of people struggling with the awful conditions created by not taking their goals into account in the plan.
In other words as long as you don't try to plan their activity. What you're describing is how more free market oriented economies work: with a central authority that has the power to force everyone to comply with its plans, but that deliberately chooses to mostly not exercise this power, in order to maintain a free market.
Central economic planning is better than the alternative (anarchy) only when addressing externalities that market forces do not address. So for instance providing a system of law, a national defense and basic scientific research. Even with all of the inefficiencies of top down planning, it still remains the only way to optimize resource allocation for the production of public goods.
> In other words as long as you don't try to plan their activity.
That's not what I said, and the two ideas are not equivalent. Agents may be allowed to accommodate locally only within certain limits, giving an appearance of free will, but still being within planned parameters and thus producing a controlled outcome. The system then will converge to a state that complies with the centrally stated goals. The key here is that individual agents are only allowed some local improvements, but they're prevented from any improvement to themselves that would damage the global optimization goals. See local search,  like for example Variable neighborhood search. 
The central authority sets very strong limits on what can and cannot be done, and enforces them strictly. It also sets up economic incentives, goals, rewards and punishments over the "free actors", like convincing workers that they need to program their lives around continuous learning and adaptation, jumping jobs every X years, and discouraging people from relying on the existing public education, healthcare and public retirement plans, which were the goals of the previous central plan, now deprecated. Anyone ''willingly'' failing to comply with the new program find themselves quickly pushed out of the system (but it was "their own fault", of course, for not "behaving rationally").
That's how the current Western world economic forces operate, and it's quite different from "mostly not exercising its power"; if that's a free market, it looks a lot like "forcing everyone to comply with its plans".
I haven't seen any evidence that this is an effective way of organizing an economic system, at least relative to the free market.
Modern western economies have grown increasingly stagnant as the level of central planning has increased since 1960. Contrary to your claim that "public education, healthcare and public retirement plans" have steadily been deprecated, the raw statistics show that government social spending in a major Western economy, the US, increased, on average, 4.8 percent per year between 1972 and 2011, after adjusting for inflation . This represents a massive shift to more central economic planning. And this shift has been associated with reduced rates of improvement in all metrics of economic development.
That's not really it either. If you have a business which is generating high profits, it will have a high market cap. You can reduce it by transferring the corporation's liquid assets to shareholders, but if you've transferred the liquid assets and you're still above the limit, transferring non-liquid assets to shareholders is clearly the wrong direction.
What you really want is for separate businesses to be separate companies. Conglomerates are inefficient. Don't expand into new markets, just give the shareholders their money. The shareholders can invest it in the new markets if they want to.
The main impediment to this is the tax code, because if you give the shareholders the money it gets taxed -- twice -- but if you spend it internally or leave it inside the corporation in an offshore subsidiary, that doesn't happen.
Wikipedia gets to the same place via a different route. A non-profit doesn't have shareholders to pay dividends to, so instead of paying dividends being discouraged by the tax code, it just isn't possible, and you get the same results. The website generates more revenue than it actually needs to operate and the rest of the money has to go somewhere, so it goes somewhere inefficient as determined by internal politics.
It's interesting that you say that on a subthread discussing something Warren Buffet said. Berkshire Hathaway became an efficient conglomerate by acquiring and operating companies that eschewed the very principle we're discussing.
Dividends are nice, but dividends would stay small if the company itself stays small (because it's constrained from growing.)
(I'm not equipped to evaluate whether or not that point is true, especially in the case of any specific company. Just trying to point out that the argument you're making here kind of begs the question.)
Look at utilities and regional banks, which are usually priced based on return on assets or operating margins.
Growth oriented companies have advantages and disadvantages. Microsoft is a great example -- they have a business that's a monopoly cash cow, but they also need to hit high growth rates to prosper. They squandered a decade trying to both grow and milk the cow, and are now growing again, while breaking and eventually losing parts of the legacy business.
That is basically what you are doing when you leave money on the table. Opportunity costs are still costs.
For about 7 years, I was collecting an effective 10% dividend while getting significant capital appreciation as well.
Likewise, as part of my diversified retirement portfolio, I have a portion in boring dividend paying stocks that generate moderate returns and don't get punished as severely during bad market conditions.