but what if we have just a lot of buyers - they just keep a constant balance level by random payments that constant level (of course, descreased a bit - just to be sure we can always pay to anyone) is the money we can invest lot of clients means high constant level which may be tens million even short-term conservative investments may be attractive with such sums what do you think?
"A lot of constant buyers" doesn't really change anything, it just complicates the analysis. The big problem is the N months between the day one sales and the time the day one revenue can be realized. Once you get past that threshold, this model is (barring the whole investment issue, which I'll cover) basically no different from a model where the customers return window is longer: it will normalize out over time. The question is is there anything in here for the seller that is beneficial enough to justify the difficulty of that first few months of dead time. I assert the answer is no.
Any investment carries risk of loss. Generally risk of loss scales with potential for return. You have to understand that a business, in this sales model, cannot consider money from sales theirs until after the N month period. That means they're investing somebody else's money with the obligation to return it later. That means they must be extremely conservative with the investment, bordering on basically not investing it at all. If the business is unable to pay their debts at the end of the term because they have lost money on bad investment choices, they will go out of business, either by their own choice or because they will be sued into oblivion by the lenders (customers). The only truly safe place to stash this money is in a bank account, which is insured against loss (although in the US the FDIC insurance limits are only up to $250k, so you won't be stashing "tens of millions" of dollars in there. I feel fairly confident assuming there is a similar limit on the "complete safe Government-backed" banks in the UK Kylotan was referring to.
Not to mention that there is nothing in a game developer's skillset that necessarily suggests they will have brilliant, mutal-fund-management-style minds and thus have any real clue how to effectively shepherd this investment. But others have harped on your misrepresentation of the value of the investment clause of this model enough, I think.
In the US you can get a high-yield savings account that approaches 1% APY. That's 0.01% after inflation, per Kylotan's 0.9% figure. On 250k that is... drumroll.... $250. And that's after a year. After two or three months (which is the most reasonable term for this agreement, I think) you're looking at about $25. You're not going to float your studio for three months on $25. Even if this number were something more realistic for covering a studio's operating cost for three months... even if you could somehow safely earn enough on these investment to get you $25,000 in three months... that's not enough because you need to keep that money invested to capitalize on all that interest gain. For the entire three months. So you still have nothing to pay your day to day costs (including payroll) with, unless the publisher, for some reason, decides to cover you for that time).
but i'm not sure this is a really big problem or at least there should be companies that could live with it
You don't seem to understand how the majority of game development companies function. They are not usually financially independent, they are generally funded by publishers to produce a game. That means it's the publisher basically footing the entire bill for the studio's day to day operation. Why on earth would they continue to pay for an extra few months of the studio running when they're getting nothing out of that studio during that time? One should not plan one's business model around the idea that everybody involves will make poor financial choices, and paying a studio to do nothing for a few months is exactly that.
publisher may pay advance money as it happens with books?
Again, here's a misunderstanding about the game development business. The publisher has already paid the developer an effective advance to make the game. The is an advance that the developer is generally contractually obligated to repay, in whole or in part, before the developer themselves gets to see any actual profit from the game. So not only would the publisher balk at the idea of funding a studio to spin its wheels for three months, the studio would as well: it increases their debt to the publisher but adds no value to the game, and thus does not increase potential sales (which would allow the developer to see a profit sooner).
don't you think it can turn more players into payers?
No, I think it means fewer actual paying customers. Customers are not altruistic in the general case, they will make the choices that benefit themselves over you the game development. If this means they have extracted sufficient fun out of a game in the N month window, they will return it to you. They will not, in the majority case, elect to let you keep their money just because you're such a nice guy who chose a business model that favored the customer exclusively.
exactly! with the hope that they appreciate it and multiply greatly
As above, this is unrealistic. This paper (and perhaps you) is analyzing a business model in a fantasy world where everybody loves everybody else and appreciates their good faith gestures and sings Kumbaya every night around the campfire.
This is not the world we live in.
and microtransactions are just shaded - they hide in short-received interest from the money a player deposits to the game
This does not make any sense to me at all; I cannot figure out what you're trying to convey. But microtransactions are not interest, nor are they investment. They are a simple, direct exchange of currency for virtual goods or perks, completed immediately. They're not hidden or shady at all, accounting-wise.
Sometimes there are some interesting accounting issues with them, for example if you exchange real money for virtual currency, and then use the virtual currency to purchase in-game items, sometimes for accounting purposes the revenue for the transaction is not realized until the latter transaction (virtual currency for virtual good). This is not unlike the way gift certificates at retail work commonly, though.