"Fixing" old code is not that easy; in fact I'd argue that its frequently easier to simply recreate from scratch.
Case-in-point: I've recently got back into C64 programming, just as fun hobby and a bit of nostalgia as it was what I started my dev career on back in the mid-80's. Now, while I have been working on a new project for awhile, but as a side-thing I've also been working on bringing back an unfinished game I did back in 1989 for which I no longer have the source code.
Basically I started with a snapshot made with a freezer cartridge back in the day, and nothing else!. Its enough, but its been way slower going than I expected even though obviously I do have the advantages of modern tools and having written the thing in the first place - albeit that "in the first place" is a time in my life 30 plus years ago that' pretty damn fuzzy!
Point being, doing small stuff like adding "trainer" type features like infinite lives or level skips isn't bad at all; but chasing down bugs and rewriting sections of code to change function is a lot more work on top of coding the thing I want to code, because you have to fully understand how everything works, where it sits in memory etc. And of course sometimes even when you do have a clear understanding of everything bugs can still be difficult to isolate and squash.
Obviously my ability ,(or lack thereof !), is the same so I think I can make this observation fairly having recent experience doing new original work on the same old tech.