Some interesting ideas. I think the cheaper barracks and initial +10 pop are pretty good.
As for plate barding armor or guard towers, that’s not something I would really want. See, there are (broadly) 2 ways to improve a civ. One is to give it more options (“flexibility”), and another is to increase the strength, utility, or cost-effectiveness of the options it already has.
I would argue that “flexibility” has never really been part of the Goths’ identity, and perhaps that they’ve intentionally been designed to be inflexible, but really good at one thing, and in fact more so than any other civ. As the ancient Greeks said, with great applicability to the powerful yet inflexible Greek phalanx, “The fox knows many things, but the hedgehog knows one big thing.” Some civs are foxes, and are good for having lots of tricks in their bag, and other civs are not as versatile, but are really good at executing one type of strategy (another example of this is Mayans, who are excellent with archers, but don’t really have other good options besides the eagle). There are more civs that are “flexible” and able to execute a wide variety of strategies at different stages in the game, and I think Goths would lose a lot of their identity if they were to become less one-dimensional. Or, to another point that has been brought up in this thread, why would you make purely generic units for which you have no bonus (e.g. Saracen or Mongol champs)? Sure, in a pinch it would be nice to have better cav, but making infantry viable for Goths is the most important thing, and if a civ tends to use units for which it has no bonus while ignoring units for which it does (an awful lot of pre-de pro games featuring Goths involved going knights or crossbows, or just slinging in team games), is a symptom of poor design/balance for that civ.