Even back in Waves 1-3, it was clearly a multi-step process. The rough formula, as reverse-engineered by a poster named ShadowJak, was 2 + Pilot Skill + 8x(Attack-2) + 8x(Agility-2) + 4.25x(Hull-3) + 4.5xShields, with an extra point (usually) for an elite talent slot and an extra point (sometimes 2 points, although not in Wave 1) for a pilot ability.
But even in Wave 1, it was clear that things were tweaked based on playtesting. By Wave 3 the formulas were consistently off-target, and most people stopped trying to use them.
And that formula wasn't a particularly good way to approach things anyway, since it makes the bad assumption that various parts of a ship's stat line are independent and have a linear effect on the ship's value. We know now that's not true: for instance, extra hit points are worth more on a high-Agility ship than a low-Agility one.
Another, more valid way to look at things is to start with the average damage a ship deals in a number of typical scenarios and the average damage the same ship is likely to take, and use those to determine a ship's value. But that requires a lot of work and a number of critically important assumptions, and I don't think FFG is actually doing it.
I used that formula, with a few additions, to make a ship cost calculator with google docs' spreadsheets. It's usually accurate enough for small & huge ships (about +/- 1.25) to get a baseline to work with. I haven't quite figured out the modifying factors of large ships, which can vary wildly from what the formula predicts. K-Wing, Punisher, Kihraxz, and Aggressors are all accurately predicted, but other stuff like the Punishing One, nd VCX-100 are pretty off the
You need to account for some factors that a statline formula wont include, such as dials, upgrades and actions. I think that comes from playtesting and looking at available upgrade combos, then adjusting the baseline costs, and playtest some more.