I agree with his thinking that the wide variety of LTE frequencies available for different operators in differernt regions is a jumbled mess. It will be hard for handset manufacturers to get the right mix of LTE frequency bands together with GSM and 3G, and therefore will likely impact global roaming. It will probably not be possible to have handsets supporting so many regions any time soon.
The result is that LTE devices, when roaming, will likely take the path of least resistance and simply camp on the 3G network.
What gets interesting is the impact this 'new reality' has on the current voice over LTE debate happening in the industry.
For the sake of argument (and simplicity), let's accept that handsets will roam to 3G networks when outside their operator's home network. How does this assumption impact an operator's approach to voice over LTE?
First and foremost, this completely eliminates CS Fallback as a potential technology to support inbound roaming. If the phone can't connect to LTE, why bother implementing CSFB, a technology which assumes an LTE connection to 'fall back' from? Just start the phone/session on 3G and move on.
Second, one of the concerns in the industry is potential 'fragmentation' in the voice over LTE market. The logic goes that if different operators implement different technologies, there would likely be a lack of interoperability.
But if we concur that handsets are unlikely to roam on LTE networks any time soon, then this really becomes a moot point.
Then, operators become free to implement a native voice over LTE solution (e.g. VoLGA, IMS Telephony) for their own subscribers. If their subscribers roam, they will roam to 3G, making the choice of LTE voice technology an internal/domestic decision.
Frankly, this is a win-win situation for LTE deployments.
Operators win because they are free to invest in a solution which matches their own situational requirements - they can decide on the voice over LTE strategy that makes sense for them. Operators can choose to start with VoLGA and move to IMS, or wait and jump right to IMS, depending on their own network evolution plans, business/investment timelines, competitive pressures and so on.
LTE vendors also win because operators will have separated their own voice over LTE decisions. Therefore, the market as a whole would progress at different rates to meet requirements in individual markets.
For some operators, there is a need to move forward with voice over LTE sooner rather than later. But if the market is forced to wait on a single IMS telephony specification, it would impact all levels of the IMS ecosystem, effectively dampening the market for years to come.
Actually this is the situation that Verizon Wireless is in right now. The company is charging ahead with LTE, but their flavor of LTE devices and their approach to LTE voice will be materially different from the GSM community for many years to come. Certainly they are targeting CDMA/LTE devices that are unique to them, and they will likely be using a non-GSM standard voice over LTE approach for the foreseeable future.
The big question is: Does it matter?
I posit no. There will be no devices which work on or with Verizon's LTE network for a long time. And if the handset manufacturers develop specific LTE frequency bands, it won't matter for other operators either.