I'd seriously recommend that you consider cutting this. Perhaps keep the lessons you've learned to retain future flexibility, but get good on one toolkit first. GUI toolkits, GUI bindings, GUI in general drowns you in details as it is, volunteering to drown in several different underlying toolkit's details may sound appealing because you may feel like you're growing your metaphorical market, but what you will almost certainly end up with is doing all toolkits badly instead of even one toolkit well, and that won't be good.
We've all heard about how the first 90% is 90% of the work, and then the remaining 10% is another 90% of the work. GUIs make that look like hopeless pie-in-the-sky optimism, where the first 10% is 90% of the work, and then the next 10% is ten times the work, and then the next 10% is another ten times the work. Trying to be cross-platform will strangle you.
However, based on my experience with some similar previous discussions, I don't expect you to immediately buy and agree with my arguments. What I'll really suggest then is to keep what I've said here in the back of your mind, and when you find you're staring down the three toolkits that rigidly require a three-way mutually contradictory way of handling rich text or something, then think back to this post and consider giving yourself permission to drop all but the best-supported and/or most popular underlying toolkit.