I'd like to add to this. There is definitely a barrier of entry with regards to setting up a full node. Unless you're living in a first world country, the bandwidth requirements alone, will outright prevent you from even setting up a full node (sync since genesis).

To maintain that also becomes a sunk cost, as there is no financial incentive to run a node, only an idealogical one. Most of the people who benefit and will benefit from Bitcoin, are the un-banked. Which you will find in 3rd world countries, that don't have ISPs that provide the data packages, to cater for the requirements of running a full node. I'm sure many would like to, but simply cannot afford it.

A user may not want to run a node at home, but rather on a digital ocean or AWS server, which they cannot afford to do either considering the bandwidth and storage costs associated with it. However, I don't think they should be excluded from participating in the network (supporting proposals, voicing their opinions, running their own wallets, writing their own applications on top of Bitcoin [which I think is extremely important]).

So I would definitely be in favour of a small node of sorts. It will present us with some interesting technical challenges along the way but it's definitely worth while looking into.

Financially incentivising nodes is a really weird area because it would allow someone to essentially automate the deployment of nodes. i.e. if a node can pay for itself 100% (even at a lesser value, it just becomes cheaper overall), you could write an application that uses an AWS API or a digital ocean API to automatically deploy 100's of nodes. Which sounds great but not if that person is malicious and wants to prevent the community adopting proposals.