How does selling Openreach actually achieve better broadband?

There are lots of news reports [BBC article] about a report a 121 MPs making up the 'British Infrastructure Group (BIG)' have written.

The report says there are 400,000 small and medium-sized companies that still did not have access to superfast broadband and that little would change unless BT and Openreach were formally separated.

What would separating the two actually achieve? What are the factors limiting the rollout of superfast broadband? Which of the limiting factors, if any, would separation solve?

As I understand things the limiting factors for superfast broadband are:

• cost - whatever the rollout technology someone will have to pay for the investment. Currently Openreach has two rollout plans:
• Commercial - this is where Openreach look at the number of  site/buildings a network upgrade will help, cost of the network upgrade and use a statistical model to determine how many sites will actually take advantage of the upgrade. If the numbers work out Openreach will invest and should make a profit.
• Non-commercial -this is when an upgrade fails to be commercial and Broadband Delivery UK funding is used to bridge the funding gap.
• technology - simplistically there are 3 types of technology for delivering broadband:
• over fibre - a fibre connection from the network directly into the building. Signal speeds are limited by losses in the fibre cable but these are very low and allow for very high speeds. The fibre links run from the exchange to each building with some passive optical splitting to reduce the amount of cables needed. Fiber delivers the fastest network speeds but requires fibre cables to be installed and this is very expensive due to the amount of physical installation work needed. In FTTC enabled areas FTTPoD is available but it's expensive due to the need to physically install fibre to the building. There are a number of FTTP trial areas where for the same cost as FTTC you get FTTP. The costs are kept the same to encourage take up of the service so installs can be done enabling Openreach to collect data on the actual time taken to do the installs. Given the need to run a dedicated fibre link (from a concentration point) costs are going to be higher then using the existing copper.
• over copper - initially available over ADSL, then FTTC (VDSL) and soon FTTdp (G.Fast). Signal speeds over copper are limited by cable resistance, inductance and capacitance. ADSL uses copper from the exchange to the building; FTTC runs fibre to a green streetside cabinet and then copper to the building; FTTdp uses fibre to the distribution point (typically the top of a pole or underground junction box) and then copper to the building. ADSL, FTTC and FTTdp deliver increasingly faster network connections due to the shorter and shorter copper cables and is pretty expensive to install due to the physical installation work. ADSL is cheapest as all that is needed is new equipment in the exchange. FTTC requires additional street cabinets with a power feed to receive a fibre signal and inject it into the existing copper street cabinets. FTTdp needs pole top (or in junction box) equipment with a power feed to receive the fibre signal and inject it.
• wireless - a mix of satellite and mobile data technologies using satellite dishes or mobile modems.
• natural monopoly - many infrastructure projects are natural monopolies. We don't have multiple electricity cables or gas cables coming to our door. Nor do we have multiple postman delivering letters. Why would we expect multiple copper or fibre lines? When the cable TV networks were introduced there were lots of cable companies but they did not have overlapping network. They had to install their own cable networks and that was expensive. There was a huge shakeout in the cable TV industry, companies went bust and investors lost their money. We now have one cable TV company after Telewest and NTL merged and bought Virgin branding. In the US the 'baby Bell's all had local monopolies an are merging back into large companies again. I really don't see the market paying for multiple copper or fibre lines again.

All this costs money and people want a cheap service. I don't see how selling off Openreach will change these limiting factors. I don't believe that the technology options will suddenly get cheaper. Personally I think the staged roll out of ADSL (just to exchanges) , FTTC (just to 90,000 or so cabinets) and FTTdp (several million poles/junction boxes) makes commercial sense and can understand why it's being done that way. Even if we wanted a full fibre rollout (to 27 million houses)  would the first recipients want to pay the high cost? FTTPoD (to the premises on Demand) is available but very expensive due to the construction costs.

There's a huge difference in cost between running fibre to 90,000 new cabinets and running fibre to 27 million homes. Just think about the amount of staff that would need to be taken on to do all the digging, installing and covering up. Also, it's far cheaper to do one area completely that it would be to do the work as orders came in. Can you imagine the disruption if in a given street, orders were made every few months and the pavements constantly dug up and put down again - once per order.

I suspect that labour force cost and availability is as much a limiting factor as anything else. Staff up a nation wide fibre roll out would need a huge labour force for a period of time (shorter period would require a bigger labour force) and once the roll out was done a much smaller labour force for maintenance and new installs would remain. A rapid roll out would be hard to staff and make many unemployed at the end. 

I simply don't see making Openreach independant as much of an improvement.