With point one, I'm in full agreement. A lot of abilities seem like they'd be great to use just based off of their description. Peregrine in another thread stated that almost any abilities that alter 'rates' whether it's success, quality gained, or durability loss are at best negligibly useful. Maker's Muse, Fulfillment, Masterpiece etc.. all of them seem like they would be really great. But in practice they don't seem to have much of an impact. Those skill are all freebies, with basically laughable requirements in the current iteration of the game, r10 in one craft, that's like 2 or 3 locals? The best skills like Hand of the Gods and Perfection they require some sort of work. Granted there a lot of abilities that require ranks and marks that are still useless and I would like to see them do something rather than nothing.
For your second point, I'm going to respectfully disagree. As I've personally seen the right gear really make the difference in synthesis under a variety of circumstances. I'll explain below how I think they come into play.
From the Lodestone Q&A on Synthesis, we know that specific recipes have certain statistical requirements. Why doesn't this seem immediately evident? Because, as best as I can figure the requirements are 'soft' meaning that there is no barrier to attempting the synthesis as long as you meet the necessary rank requirement. At which point you can attempt the recipe in your underwear and with a weathered tool if you'd like. But if you're not within a certain threshold for a particular statistic you're going to have a heck of a time overcoming a certain amount of failures.
To me this became very evident on my ALC when I tried to break clusters at r38. It was absolutely miserable until I played with my stats through gear, and upped my base Control and started eatting Seafood Dishes, and at that point the synthesis became much more manageable. My success rate swung 20% in my favor, and has consistently stayed that way.
So what I think is happening is that most recipes just utilise these soft thresholds for meting out a certain swing in success rates. And people tend to think that stacking Craftsmanship, or M. Craftsmanship should linearly increase the success rate all the way to 100%. Well no, that's not how it works for STR or INT to damage, and people accept that, why then should the game use that approach for crafting? (One exception is for HQing I definitely stack Control as much as I can, while still meeting what I percieve to be the threshold for Craftsmanship or M. Craftsmanship for the recipe, because Control does seem to delay the appearance of sparks).
Also as your parenthetical note there, I don't believe that STR (or any of the 6 primary stats) has anything at all to do with success rates. In another article on Lodestone the devs stated that the effect of statistics on crafting was limited to HQ rates.