I know there already is all of stuff that we need to research in the game but I really would like for complete ships to be needed to be researched in the same way we need to research components. Building new ships in reality are a pretty research intensive thing to do even if most the individual components are more or less known or common. Everything always need some sort of special treatment to work in practice.
There could be a template ship type and then you can do variation of that same ship... so variations should then be a separate thing. The template is the basic ship hull with a basic sett of components you then can do variation of that template which are much cheaper to research, similar to retooling for a similar ship.
When you update a template to a new ship type that are similar to the old template is also is cheaper but not nearly as cheap as just doing a variant ship type.
In this I think that components probably need to be weighted not only in cost but also in size and/or type depending on if the ship is variant or you need to produce an entire new template. You also would then replace the retooling system with this system as well. A shipyard retooled for a certain template could also build all the variants.
It would be a nice opportunity to make ships more into a modular lego system rather than the more free form we have today where anything goes everywhere. In reality I'm sure we could not just put any component in any place on a ship without some consequence. So making some changes to fit such a system into the design would be nice as well. It would put a few more restrictions as some components would compete for the same space on the ship while others won't. There definitely should be a large difference if a component need to be placed closer to the edge of a ship or not or if it matter more or less to some components, this then could also have an impact on how different components are exposed to combat damage as well rather than just being based on size. Sensors for example usually are pretty small but often will have to be placed in compromised locations so should be much more likely damaged when just looking at pure size. Compared with an engineering section that likely are located much deeper inside a ship and thus less likely to be hit.
I also think that the research cost should not scale linear by size so a small ship should be allot more expensive than a larger ship seen to their mass. A small fighter should be relatively expensive to design versus a 100kt carrier seen to their respective size.
In any way... we then would have to research each ship design and all its variants and then we retool the shipyards to build them. This would give us a more "realistic" ship design procedure... in general I don't think the retooling cost is nearly enough of a cost to represent the investment in building new ships.
Obviously not a small change so nothing to take lightly...