You seem to have everything backwards, which makes me question whether you read the actual comment as opposed to just attempting to guess at things from the snippet :(.

What you said would make some sense if the client connected to someone in the middle who then connected to the server. Instead, the client connects to the server, but when data comes back it is forwarded to someone in the "middle" (really, off to the side) who then processes it and returns results to forward back to the server.

In essence, the device is in the middle of a connection from the developer to the service, as opposed to the developer being in the middle of a connection from the device to the service. In such a configuration you are not getting any of the benefits you are talking about, as if the device isn't online it won't be able to maintain the connection on behalf of the developer's server.

(Also, I must point out the normal situation with such messaging apps on iOS is that you get them from the service itself. In addition to being less common on Android, it is thereby interesting to point out to people who might not realize the app is implemented in this weird way--some people who even analyzed it assumed this was a side-channel for analytics--so they can make the right call for themselves.)

This is very standard with web scraping implementations. Honestly, if you are reversing someone's protocol or scraping their web site, you are stupid if you put the implementation on the client because it will break all the time. Instead you put it on a server where you can fix breaks instantly. The server usually delays/shows stale data while things are broken as well, then does what it can to replay once fixed.

On the other hand putting things on the server means that you can get cut (by IP) and also brings the need of dealing with scaling problems.

You can always scale easier on the server platform, as well as spin up new instances with new IPs on any VM platform.

Handling it on the server allows you some control over the chaos. Not so much when all the logic is on your mobile app.

You aren't suggesting that it scales easier on the server than on the mobile client, are you? Nothing would scale better for this app than putting the logic on the client -- it's a customer's phone communication with an Apple server. This leaves nothing for him to scale, anywhere, except for engineering to push out updates.

I wouldn't say you're scaling properly if you have millions of mobile clients failing to make the proper calls to a service because the protocol has been changed. I'd say its damn near irresponsible to do that. It's not scaling if you're just throwing packets against the wall when the protocol or service unexpectedly changes.

The impression I got was that actual requests and responses are still made by the client.

>stupid if you put the implementation on the client because it will break all the time

Makes some sense to do Apk downloads for this unusual implementation of it allows a fix push

Thats how jabber works afaik having a network of domains and each domain can have a set of transports to connect to various non-jabber networks, this might as well be a proprietary implementation of transport and private service.

Anytime something might smell like a certain lawsuit you bet you will be on the short end of the stick - with corporations.1

AFAIK there is still no way for a messaging app on iOS to reliably run 24/7. As a result, most messaging apps use their own server for new message notifications. If the app is not running, you get a notification and you can open the app to read the new messages. That of course means a MITM-like server …