For transportation there are driveways, small roads, highways, trains, airports, etc. They all compliment each other. Same will happen with DLT.
Just as there are networks of networks making the internet. There will become ledgers of ledgers for a variety of data sets. Each layer providing different features. I don’t think things will evolve as simple as public / private chain. But layers of private chains communicating with layers of public chains.
Right now businesses keep their data on central servers, and have to pay for ongoing upkeep, security, redundancy, insurance, etc - and is more susceptible to human error & malicious intent.
Some of this data also needs to be available to third parties / customers / clients / etc. This is where a public gateway in necessary. Instead of giving the 3rd party access to a data set within the central network, that data can be encrypted and put into an additional layer of which the individual client has the keys to. - Much like your current credentials grants access to your existing accounts.
Lastly, these individuals will need the means to share this data with additional parties. EG. You request your bank statement, which is encrypted and published to a public address you have keys to, from the banks private chain. You then forward these statements to your accountant, who then logs them within the office private chain, completes your tax prep, and again encrypts and publishes them to a government address on public chain. And once again, they store that data in their private chain.
And even within this quick example there can be a multitude of other “things” happening, such as zero knowledge proofs, onion routing, atomic swapping, multi signatory, aggregated transactions, non fungible asset transfer, etc.
We’ve barely begun to scratch the surface of what is possible.
Long story short, NEM needs to remain competitive for both public and private chain applications, as well as remain open minded and adaptive.