I'd make a slurry of the coco and test pH .. make sure it was properly ph-balanced. Some coco is not processed properly when you get it and can kill plants or make them sickly. Seen it happen. Seen it happen to others.
if coco is fine...
Use an apprpriate size pot for seedlings, despite the bro science that says otherwise because 'autoflowers are magically different than all other plants.'
Potting up does not shock a plant unless you purposefully treat it roughly. Potting up will result in an ideal rootzone and eliminates the needless addition of difficult watering habits for a tiny plant in a huge pot. a healthier rootzone with less risk of funky microbial growth from stagnating water.
So, it's been irrigated to start. Compress a small 1/2" deep hole. As opposed to digging, it creates a safe small compressed spot for the seed to rest without fear of it falling deeper into medium. Drop the seed in, preferably not upside down but even that only slows it down a bit. Gently cover from sides and tamp down - not too tight. To avoid a '2nd' irrigation before they sprout, i use a humidity dome. I take the dome off when they sprout.
Based on sprout times, adjust depth you plant the seed. You are too shallow if the shells remain on the sprout. Minimize that shell-head effect while getting the shortest sprout times possible.
Keep it simple. From what i read the difference between direct and paper towel is so minimal it won't matter when dealing with a handful of plants in a home garden. One thing that isn't an argument is that the paper towel method is not as simple and requires more hands on effort while exposing a delicate tap root to light - not ideal. What's fun is that is far more 'shocking' to the plant than potting up, but "that's" okay, lol. inconsistent nature of bro science...
regardless, you will have a slight learning curve for each -- like choosing depth for direct method. these issues are temporary and irrelevant in the long-term. You'll figure it out either way and have ~100% success with viable seed. However, you will run into inviable see occasionally and that's mostly about how old the seed is -- even that will vary by genetics as far as when they start to fail. I have some 3-4 year old homebrew seeds that still pop near-100% and i don't store them in the 'best' way... just a drawer and a baggy to give a point of reference.
Is it your fault? Is it the coco? Is the seed just old and inviable? That's for you to investigate.
--------
one other thing because the way you mention how you water is concerning.
Never partially water. All this does is cause a poor drying pattern that trains superfical roots.
1) fully water. * with coco 10% runoff in addition.
2) wait for appropriate dryback and repeat
In a soilless medium, like coco, you should fertilize everytime with a well-balanced formula. You provide 100% of the nuets. You want a well-balanced formula around the roots at all times. It's not about proportion of use in the plant but about ratios that make each easily available to the plant through the roots. The 10% runoff eliminates buildup of nutrients in teh medium. Any issue you see if fixed through adjust the formula and no worries about buildup being a cause. Easier diagnosis is a benefit.
With seedlings i wouldn't let the surface dry as far as the '2nd' irrigation is concerned, but after that there's enough roots to make use of a less risky wet-dry cycle. With coco when it starsts to turn color on top, usually a good point to re-irrigate. You can go by weight / weight loss, too. If you re-irrigate at same weight loss, it will require the same volume of water.
This should all be a reaction... use as much as it takes to get 10% runoff or more... wait for observable weight loss or drying up top to trigger next irrigation. Don't choose either in some inductive manner. That is the wrong way to do it, objectively speaking.
i'd look up the end result of jacks 321 / cropsalt / masterblend / mega crop hydro setups... mimic their elemental ppm ratios and start at around 600-700ppm overall concentration (apps and websites can calculate this from gauranteed labels and dosage instructions.. some can give dosage for specific ppm targets etc)
There's a reason all these companies use a similar ratio once fully mixed. It's all based on the same existing knoweldge instead of trying to re-invent the wheel with some esoteric bro-science formula tha tonly works for 50% of plants, lol. A good formula wil work well on vast majority of plants. if a formula consistently results in leaf symptoms, that's telling you it's not a great formula to use.
keep track of elemental ppm over time and you'll see what causes a deficiency or toxicity and know to avoid it. Whether i'm runnign 2 strains or 8, this formula (used by multiple sources) works on vast majority of genetic variety. This will shorten the learning curve.
This is part of he super power of soilless. Easier diagnosis, re-formulation based on observance for better, consistent results. Optimal levels of nutes around roots creates fastest growth.
you may learn that you need a slightly different formula for flower than vege. Less N is usually the adjustment. Too many variables and choices you make could impact exactly what you need to do. I drop N about 20-30% in 2 steps as i enter flower. How much you fed leading up to that can make a big difference.
---
check out guids on cocoforcannabis.com -- coco is not necessary. any soilless option is similar. Their guides and articles are bar-none better than all the other marijana sites out there that delve into magic far more often and oblivious about it. There's less anecdote, more rational though invovled. They are honest when they extrapolate something that may or may not be true, for example, wherease the other sites state unsupported things as fact. It's a more honest and rational source for info.