O.K., and thanks. The explanation makes sense and fills in that gap.

Let me then ask this, if you or anyone else may have the answer. I tried to graft the section for an installed, known to be working prior version extension -- from one Opera ver. 99 to another Opera ver. 99 on a different machine. This will be a crude, inexact representation, but close enough, and hopefully the idea will come across:

This (example): 

in Opera, under \Data\Profile\Data

\Extensions
\bdallnkjchemfeicfdneeaidongbpdnj
\1.1_0
\metadata

\kgodbpmaebgibhnnejlcmcfpfenfmfo
 \0.1.1_0
  \data
   \icons
    \interface
     \resource
      \vendor
       \codemirror
        \lib
        \mode
        \sql
        \sql
        \lib
        \chrome
        \metadata

I realize that these extensions are not explicitly identified (until you look in the Manifest.json under 'Name:' for each), otherwise they are just getting some code ID string as seen above.

But the graft wasn't seen, and did not work. My speculation was that when the extension got installed, something essential also went into some Opera registry file(s) for them ?

Another question was whether there was something in these sections to click on, that might install what is in one of these older, transferred sections ?

I'm just trying to extrapolate from some brute force type hacks that I think I may have managed to do with other browsers in the past -- although their designs and structures were different. Sometimes such a thing might work, provided one had some basic grasp of the rules.

All very well, but as I wrote I am running v. 99, not Opera One, so that does not help me. I still need the previous (now withdrawn) extension, so I'll be dependent on whoever can supply it, or wherever I might still be able to find and install it from.