This isn't something I've thought terribly deeply about, and it's kind of a crazy idea, but it's something I was musing on.

How would things change if you used App/10 as a multiplier on all Glory earned? So that normal App of 10 would get normal glory, a lower App would get less, and of course a higher App would get more.

For the general part, I feel like this would be a strong incentive to put some points in App. And I feel like you still wouldn't want to max out App at the expense of everything else. App becomes kind of like a 10,000 Glory bet - if you make it to that much "natural" glory, then the extra glory will be enough to make up for the points spent on your App score. I feel like most characters don't make it that high, but it's not that unlikely. And while it would be extra math, I don't think it would be that bad because you'd only do once per year, when you're already adding up all your glory anyway.

On the other hand, it creates the potential for an old character with a high App to live a lot longer, and it potentially contributes a lot to glory inflation.

Thoughts?

I think it might be a bit too much at high APP, and it adds another layer of calculation. Not that bad if you automatize it on an excel sheet, but still... A defeated knight is a defeated knight, IMHO.

However, my preference is moving towards simply giving a 'Courtly Skill' bonus of (APP-10). So someone with APP 18 would get +8 to Courtesy, Flirting, etc, making him a very superior courtier. It would also accomplish a bit of what you are thinking, in such at that Glory skill bonus would be effectively diminished by low APP: Two knights of similar glory, the people would assume that the more handsome one is more glorious and worthy of attention.

EDIT:
There is also the fact that by making it APP/10 as a multiplier, someone with APP 15 will gain three times as much Glory as someone with APP 5. This becomes a huge difference quite quickly: By the time APP 5 guy has gained 3000 Glory during his career, APP 15 guy has gained 9000, assuming that they'd gain Glory at about the same rate (safe enough assumption since combat glory is not that much anymore, and most of the glory comes from completing quests and battles). Point being, the APP 15 guy would have 6 extra Glory Bonus Points, so we could end up in a situation where we have:
APP 5, Damage 6d6, Sword 20.
APP 15, Damage 5d6, Sword 26.
I would bet on the high APP guy to win, meaning that not only is he handsome and glorious, he has become better at fighting than APP 5 guy, too. He will be getting to the Round Table much more easily than the APP 5 guy and whilst that is moderately understandable, the edge is way too much.

Now if you were to use (10+APP)/20 as a multiplier (or to put it in another way, each point of APP below 10, -5% to Glory, and each point above 10, +5%), you'd get something between 0.75 (APP 5) and 1.25 (APP 15), which would be much more agreeable, although a bit more of a pain to calculate.

In our campaign, APP is rolled randomly (although you can increase it by spending other Attribute points, but not vice versa). That seems to work fine. I am still tempted to just use the (APP-10) as a modifier to skills where presentation counts.