Here we get into the opinion realm,, hasn’t stopped me before however,,
The most pragmatic problem is historically data has been managed by the IT folks, and voice by a telcom (facilities)manager. There are often parochial turf wars, and rarely any corporate technology road map. Accordingly, both groups pursue their interests with little (no?) joint planning or consultation.
Next, the advent of “IP” phone systems has been bally hoo’ed for quite a few years, but only recently has there been legitimate systems introduced into the market. To some extent this has been a technology factor, as voice systems require immediate action (interrupt state driven) and the processors (in the pc world)have lacked the through-put (horsepower) to handle a real telephone application.
Combined with this, circuit switched (traditional) telephone systems are inexpensive. Any of them can be hooked up to an IP connection with standard analog or ISDN trunks. This has slowed the market perception of IP based systems as being a true technology revolution.
The real savings on circuits are when there is distance involved. If there are toll (long distrnace) charges incurred, then there is an economic decision that can be made. In the local network or enterprise network, the savings are nil. The local calling area is often essentially free and within the enterprise the cabling and wiring is in place. So making a change over in that environment is more difficult to justify.
Now the absolute greatest cost savings exist for multi-national companies. But in that realm, many countries are arbitrarilly preventing IP as a voice alternative. In many of those cases, the local telephone company (PT&T) is government owned, so they stand to loose a m,aajor revenue source if they are bypassed. This is changing, but at a slower (methinks) rate than here in the good old US of A. This is where I think IP transmission may first become viewed as a legitimate alternative.
Penultimately, there is the issue of station sets and conveniences. Over the past twenty years traditional switched circuit system station sets have evolved to offer a plethora of features (bells and whistles) that the users would sorely miss. Even the new IP systems have yet to match the conveniences (and status symbols) that they currently enjoy. Unless the new technology can match or beat what they already have, they will drag their feet in supporting any radical changes.
Almost last, the competition between the carriers has resulted in customers being coerced into signing multi-year facilities contracts to get major savings. Here is the Great Southwest, US West (Qwest) has signed business customers to ten year contracts for T-1 access, to freeze out the CLEC competition. To get any significant price break, most carriers get at minimum a 36 month contract. This has occurred at the same time as the IP technology has really become attractive due to the equipment (systems) changes. True one could cut over, but they would still be liable for the unused part of their facilities for the duration of the contract.
And in summary, I think the market itself has not yet been convinced of the merits of IP networking. I think that it is still viewed (to some extent) as a nerds-ville solution, and major communications managers are not going to put their (expletive deleted) on the line and do something radically different until it is well accepted. It is a little like the cliche that “no one ever got fired for buying IBM” and there is a bit of cynical truth in that statement.
Don’t know if this off the cuff rambling addresses all of your questions, but it sure fills up the space..