It wouldn't be baiting less experienced players, because it would effectively place all GCD heals on mostly even footing with OGCD heals. The only difference would be the MP costs, but that's negligible in the current way that MP is handled, for the most part.

The bigger issue is that it's a cureall solution not unlike conforming as many jobs as possible to the 2 minute burst meta. I believe the intention behind the suggestion is to make GCD healing not feel like a punishment and allow players to diversify their casting focus to support tools comfortably. On a certain level, this would work as using barriers as a barrier healer would now be something you actually do, but it creates far more issues. First, it creates a scenario where you feel obligated to build 5 stacks before each 2 minute burst window regardless of whether the healing is needed or not. Since it sounds like the suggestion is automated, and you can reserve stacks for later, this exacerbates this issue as you'd want to blast through 5 heals right before the buff window to turn those refund stacks into DPS gains. Outside of that, it otherwise eliminates skill expression as now you'll naturally hit mostly optimized DPS values without even trying. It almost doesn't matter what you cast at that point.

DPS refund is the way to make GCD healing more valuable and create more opportunities for healers to actively choose healing over DPS, but it should be executed with more thought in mind--that it's something you have to actively manage with resources, MP, cooldowns, or any combination of the like in order to perform successfully. Heals that fall out of that range--that do not refund your DPS, should be stronger and more effective panic buttons because that's essentially what they are. You can drop your optimization for safety, in other words.