Oneye Power (and crew) mangement
Compartmentalize heavily (IMO), meaning no more than 1 reactor per compartment. But don't use copy-paste modules, that's will always be sub-optimal compared to a purpose-built compartment.
Test out each of your compartments with autofire, and make sure that you're not wasting anything. Every reactor should run out of power, every factory should be empty/nearly empty, and every crewman should be busy. There are exceptions of course, but generally you should aim for 100% power/ammo/crew usage.
I like to cut lots and lots of corners. Most battles are won in the first 30 seconds, so prioritize burst value over sustained value. You don't need enough power to keep your weapons running all the time, or to keep your shields up indefinitely, or to keep your engines firing 100% of the time. Make sure to test all your compartments as well, so you can trim out any excess crew. Efficiency is key to keeping cost down.
Of course, sometimes you absolutely don't want to cut corners. Ion beams should have 100% uptime. Shields covered by lots of armor need a full reactor, or at least 1/2 of a reactor. Engines on a kiter cannot afford to lose power. But generally, cutting a few corners will help you get more value out of your money.
I'd have to know what the meta is to answer this one! Generally the key to a good ship is to make it strongest against the meta, but flexible enough that it can fight other ship types too.
Practice, practice, practice - get a feel for your ship. Know what each type of command does and when it's most useful. Set hotkeys, and focus fire. Use autofire to get crew moving, and use hold fire to avoid wasting stored ammo on wreckage. Have a battle plan for every fight you get into, but don't be afraid to adjust it on the fly.
No idea! I never use more than 1 ship. But the most effective swarm designs I see are usually either ion swarmers or cannon clingers.
Ions are the backbone of your ship. Their job is to survive after all your other weapons are dead and to keep putting out consistent damage. Most of the time, ions are best used as a central weapon, hidden behind layers of armor, with a shield or 2 to block poke damage. Ions are generally weak vs shields because they can't focus fire, but if you have 3 or 4 of them in the center of your ship they can deal effective damage through shields without support. Ions are probably the best kiting weapon currently; if you use them for this make sure you have plenty of shields and PD.
EB are the super-powered shotgun of space battles. If you're fighting at close range, get electrobolts, unless you're 100% sure the enemy won't have shields. Protect them well, because EB get more and more valuable as the fight drags on and enemy shields get time to regen. Laser+EB and cannon+EB are the two most classic combinations. Ion+EB usually doesn't work that well, because EB need to be exposed, completely nullifying the ion's biggest advantage. But if you can protect both ions+EB well, it makes for a powerful combination. Railguns+EB suffer from the same issue as ions+EB, as well as the huge range discrepancy. Missiles+EB are just plain bad 99% of the time.
Personally, I use shields a lot less than most other designers. IMO shields should only be used when you have no other way to cover something, not as a primary defense (because they'll get ripped to pieces by electrobolts). Often I see ships from new-ish players that surround their entire sides and rear with shields - this is very bad. IMO armor is almost always better than shields as long as it's not impeding your own weapon fire.
Use shields where armor won't fit, where you can juice lots of value out of regen (e.g. covering armor), or when you're planning to kite outside of enemy EB range. Otherwise, armor is usually a better choice.
Don't get discouraged too easily! If you're trying something new, you will fail. A lot. But just because your first iteration is bad doesn't mean the entire concept is. Figure out the good parts and the bad parts of your new design - use the good parts, scrap the bad parts.
If you're struggling to think of something new in the first place, look at whatever the meta is. Then think "what's the exact opposite of the meta" and go from there.