Basically this exactly.
Getting the data just to look at comments is an API request.
Where should you get that data from?
If you want it instantly and from a trusted source then you should be delegating a tiny amount of RCs to a node you trust as payment for this service. Once the delegation is setup you probably won't even ever have to think about it again is it recharges itself and probably won't ever run out.
If you don't need the data quickly and accuracy/security isn't a high priority it could be downloaded from some other random user or perhaps even from a dynamically streamed P2P torrent. We are going to see some pretty crazy stuff over the next ten years as this stuff actually scales up and people realize that bandwidth isn't actually free. Someone has to pay for it no matter what. WEB2 dictates the centralized agent needs to pay for it... but WEB3 has no centralized agent which puts us in the perfect position to undercut WEB2 and beat them at their own game.