From my experience, large codebase refactorings are almost never a priority until the deadline is just right there. 10 years for many companies mean the last 6 months, when everyone else has already done a lot of work on the same problems as you will encounter, lots of issues have been already solved, and you can probably find people with the python 3 skill. And last but not least, you will find approvals from management side "this year's the end of support, it must be done, otherwise we are out of security patches". So, essentially yes, they gave a lot of time for enterprise companies to let others do the dirty work :)
That's true. I worked at an enterprisey shop that delayed upgrading from Java 6 before its EOL. By the time they started, the EOL date had passed a few months prior and the upgrade itself was a 1.5 year effort. Some orgs can be pushed, others have to be shoved.
