In the past, Sony's divisions didn't cooperate to the best of Sony's interests overall. Components would be sold for profit between divisions rather than the lower production price giving Sony a pricing advantage, and Sony Pictures weren't on board creating a movie service as I understand it.
That just shifts where the profit is made. And depending on demand, could entail more volatility (risk) if you sell components to another division at cost or at a loss.
For example, Sony CMOS sensors are in high demand currently. There's a guaranteed market for them that is very profitable.
Sony cell phones, however, aren't in as high a demand, and prices are under heavy pressure. Selling the sensors to the phone division at a loss, at cost, or at discount "could" allow them to price more competitively but doesn't guarantee that they'll perform any better in the market. And if they can't generate as least a greater profit margin from the sale of the phone than the CMOS sensor would have gotten on the market, it would be a loss for Sony overall. And that's assuming the phone would even get sold.
In short, in that example. It's obviously in the best interest of the phone division to get the sensors at cost or at a discount. It not in the best interest of the devices division and may not be in the best interest of the Sony Corporation as a whole with regards to profit generation.
If they're measured against their individual profitability, it'll be in the division's best interests to screw over other divisions if its better for their bottom line. If Sony Imaging can sell cameras to Sony mobile and SCE (in Morpheus final release or whatever) for a crazy mark-up because the market has high demands, they will do, rather than operating a cohesive competitive advantage across the whole organisation.
There is less and less overlap between the divisions as time goes on. You don't, for example, have a division with PCs that might be pulling sensors from your devices division, competing with you TV division for LCD panels, etc. No division producing LCD panels anymore. Same goes for multiple divisions attempting to get Cell processors.
It could also be argued that with them being whole owned subsidiaries with their own accountability, that it could drive down prices of components for other wholly owned subsidiaries. As they cannot hide costs overruns through a multitude of divisions and pricing will be more inline with what the market has to offer versus a product that may in the past have been monopolized by the various divisions within Sony leading to a cost per unit that was potentially higher than market competitors.
This potentially leads to more agile divisions that can respond to market changes more quickly. But it does come at the cost of potentially more duplicated upper management. Management will be more targetted within each division, but won't be shared between divisions. It has both good and bad points. But in this case it should be good.
And as DSoup already touched on. It allows them to more easily gauge each divisions performance, profitability (or lack thereof), and sustainability. They've identified which divisions to focus on (Growth Drivers), which divisions are safe (Stable Profit Generators) and which ones are at high risk (TV and Mobile Communications). Once they are wholly owned subsidiaries not only will it be easier to identify which "bin" each subsidiary belongs to, it'll also be easier to potentially address problems or determine if the subsidiary needs to be sold off.
Most importantly, if they have to be jettisoned, it'll be a far easier process without requiring "restructuring fees" or or "exit fees" (like the currently recurring fees for exiting the PC market). This becomes very important when looking at the TV division and especially the mobile communications division. If they are smart (and I believe they are), they've already come to the conclusion that there is a high likelihood of having to exit the smartphone market, but it isn't guaranteed that they will have to do so. The process of spinning off divisions into wholly owned subsidiaries isn't without cost. However, once it is a wholly owned subsidiary, the costs of divestiture shouldn't be more than the recurring costs associated with leaving a market segment with a division within your corporation.
So, by 2018, they'll have a better idea of what they'll need to do with the mobile communications unit. By then it should be a wholly owned subsidiary. If they then decide to exit that market, the costs should be less. If they remain in the market, the subsidiary should be more agile at responding to changing market conditions. It also somewhat isolates Sony corporation from the financial obligations (liabilities, bankruptcy, creditors, etc.) of the subsidiary.
Overall it's a good thing that they are doing this. However, you are correct that a subsidiary can and sometimes do become competitors to other subsidiaries owned by the same parent company. For example, there's a technology holding company (can't remember the name) that owns multiple competing PC graphics card production companies. Sony won't have anything that drastic. But there are likely to be some areas of indirect competition. But that isn't necessarily a bad thing as MrFox pointed out with the PS Camera. It would have been a waste if Sony Devices had to provide sensors for it at a price that was significantly lower than what the market price for those sensors are. And it would have been a waste if the division making the PS Camera had to pay anywhere near the market cost of the Sony manufactured sensors when it didn't require something that good.
Regards,
SB
. Business units' priorities will still be set by HQ and based on what Sony said today, it's no more or less fractured than the existing organisation. I don't think Sony have any business units in direct competition with each other (other than competing for consumer money) but there's no clear predatory cannibalisation like we saw Apple's iPad eat into (or slow) Mac sales for a few years.