Relative Level Instead of Objective Level

I have been drowning myself in design theory of late, and I think I came across a diamond in the rough. I guess you will have to be the judge of that.

Fact: Many many P&P RPG players enjoy character advancement. "Leveling up," if you will. Being better at challenges previously encountered, after enough training/time/exposure/experience points. They want to feel the progression. It allows them to speak confidently about what their character can do when roleplaying.

Fact: Leveling up is the #1 cause of very annoying number inflation in P&P RPGs. Dragons need high attack bonuses and armor class rating because they are fighting high level PCs, whereas goblins can have nice, small, understandable numbers because they fight low level PCs who also have nice, small, understandable stats and bonuses. If this were not the case, fighting a goblin and fighting a dragon would be equally challenging. That doesn't make sense.

(Potentially new and valuable) Fact: The objective stats of the dragons and goblins don't matter. What matters is only the relative stats of the goblins and the PCs, or the dragons and the PCs. Is the dragon a higher level than the PCs? Are the goblins a higher level than the PCs? Lower?

How to apply the fact:

• Player characters record their level simply as a number from 1 to (whatever you want as a cap). Leveling up can work however you want it to.
• When the PCs get in a fight with a monster, the higher level combatant(s) get a bonus to their rolls, and the lower level combatant(s) get a penalty to their rolls. This can be a static and constant +/- 1, or it could scale with the difference in level. Your choice.
• Equal levels across the board means no adjustments.
• You don't have to adjust most other stats for level. Ever. 

So when the level 1 PCs first meet the goblins, who are also level 1, there isn't any extra math to do.

Then, after the goblins, the level 1 PCs run into an owlbear, level 3. Uh oh, the owlbear gets a +1 to its attack rolls, saves, etc. etc. AND, the PCs get -1 to attack rolls, saves, etc. etc. It proves too hard and the PCs flee successfully.

A few sessions later, the same group of PCs, now level 4, go find this owlbear to get revenge. Well, now the PCs benefit from a +1 to rolls, and the owlbear suffers the -1.

That net change of +/- 2 could mean the difference between expecting success and being doomed to defeat, depending on your system's math. If it doesn't, make the swing big enough to cause the difference...or make the math of your system more bounded.

So player characters do get better, but not because of a big number on their sheet. They get better because they keep adding more and more monsters to the category that grants them bonuses while fighting.

Thoughts?