Article seems pretty flawed. Relevance is a vague metric, and the author relies pretty heavily on data related to government site visitation, which seems subject to bias toward certain types of users.
Market share is likely still incredibly low, but Firefox’s relevance should be spiking right now due to Google’s shenanigans with Chromium. The fact that like 90% of revenue for its for-profit wing is from Google is still troubling.
Any alternative views out there?
Some non-HTTP(S) Internet stuff:
Email is transferred to its destination (where, sure it might be accessed through a Web UI). Even where things like Slack are used internally, email usage between organisations is still extensive, due to effectively being a federated lowest-common-denominator system that’s not completely at the mercy of a single vendor.
VoIP, which increases underlies telephony networks, uses things like SIP, RTP and RTCP - even if, again, it might be accessed via a Web UI, it doesn’t have to be, and there are dedicated clients.
SSH is widely used for remote system administration. SFTP, built on top of SSH, is used to transfer sensitive data, e.g. (in the US) medical records covered by HIPAA.
SNMP is used for network device management, sometimes doing so via the Internet.
Don’t confuse certain end-user applications with the Internet more generally.
The original comment, was the claim that the internet is doing a lot better than the web.
In that context, the fact that literally every single one of those services is primarily accessed and managed through the web, makes that claim that the web hasn’t succeeded look a little ridiculous.