..
simply reinventing multibit but with a usage charge wont really take off, lets hope thats not all your conception is
You are right and (hopefully) wrong.
People that want free, well they can use one of the many free wallets out there.
What I see all day every day is the need for support. Businesses run things, but when they go wrong, they don't want their staff to be on google trying to find answers. They need: call this number > get a ticket > get a resolution > close ticket.
What you also have to do is take a look at the world from outside the day to day Bitcoin bubble.
You would have no trouble running an SPV with a full node attached, with encryption enabled. Would an ordinary worker at a utility company have that level of skill? And why should they have to learn it? They are there to do a job at the company they work for, and they are expert at that job. They don't want to have to become expert at running something their employer wants them to use.
And think about running full nodes from an enterprise point of view. An organisation of 1,000 people decides they want to run Bitcoin and they need 200 full nodes. That, like it or not, will slow down their already congested networks. Bitcoin wouldn't last a week once deployed on production environment. Scale that up to 100,000 strong business needing 1,000 full nodes and you can pretty much hear the CTO laughing you out the door.
The financial investment should be compared to complex "Point of Sale" and data archiving for the finance division of the company. Most of these
companies will only run a few nodes to support decentralization {out of principle, if they understand the importance} but the rest would be API's.
Previously many of these companies dumped millions into complex mainframes & networks, so this should not be a real issue. I know our IT
guys are running a few nodes, just to experiment and to learn about the technology.
... The rest are hosted by the Finance division.