The ability to program coupled with ability to design games (both in the
https://docs.google.com/document/d/1iNSQIyNpVGHeak6isbP6AHdHD50gs8MNXF1GCf08efg/pub?embedded=true sense and "
tell me about
game theory" sense) is the best. If you can at all start from that then I would strongly consider it, even if your lesser talented mate has made something fancy while you are still getting to grips with some rudiments of programming.
There are engines that game devs will use to prototype mechanics that they will later build from scratch, though that they later build them from scratch says it is about as useful as the cardboard model that some engineering types might start out with.
There are those that will twist things. Ao Oni used RPG maker, but was anything but an RPG. While that might be an example of overcoming restrictions, or maybe playing to them, then said restrictions will always be there. I don't know how true "you will always think in terms of your first game making tool" is when compared to languages both programming and spoken, where it is rather notable, but I would not bet against it. Restrictions and limitations are why most game devs will eschew them and be something you constantly butt heads with, and have been the case in every one I have ever tried, and imagine it always will be until we get something like natural language programming.
Going back to the opening bit then you may also want to be flexing different aspects of game development muscle -- if I have to build an engine or work within a junk one then my track building for racing might be rather hampered vs finding a game with mod making tools.
Said mods if they are close to a base engine might also be a thing; if you picked say unreal for your engine then you could go from scratch, or find a game that uses a largely stock engine and mod that to have an easier in as you have working variables you can start from.
That would also speak to the other things of what do you want to make -- Unity might do well for exploration and 90s PC platforming but probably going to be harder to do a racing game or complicated RPG that RPG maker might do.
At the same time it is noted in programming tuition, and other practical skills type affairs, that the best motivation is a project that you will do. If it turns out that the existing stuff is what will do that for you then yeah. That does also mean you can grab
https://trilarion.github.io/opensourcegames/ or
https://osgameclones.com/ and play with that.
