Yes. It could be further clarified that bots and scanners are secondary when software tries to mimic human behavior. So your high-level distinction is still worth it.
The huge difference is that websites have a dual role.
- give information
- real information
While for web services there is no presentation concept. You will find the same difference in your expression languages: while HTML cumulates both informational tags and presentation directives, xml is just information identification, organization, transformation and organization.
Historically, XML followed HTML, when people found out that there were more efficient ways to access the information provided by websites than just copying their unconfigured html pages, more or less imitating people; while at the same time everyone knew that neither CORBA nor RPC could satisfy the need for B2B communications because of the impossibility of routing through the WAN.
Consequently, SOAP, all OASIS standards and only the latest REST services, still prefer devices to be too "light" to host full-fledged SOAP packages.
source share