The current dreadnought castings are much better than the first run, this is true. There is also a another factor to consider - lose the resin, lose the dreadnoughts. It is not as simple as just switching the material, as casting large pieces in metal brings its own issues (we have been down this road before).
Now, we know a lot more about resin casting than we did even a year ago, and you can see this in (for example) our Victory at Sea range, where we are running at close to 0% failure rate. The hybrid Battlestation is another example, which you will see in February.
The difference is in the model's design, and it is very difficult to turn SFU ship designs into resin-friendly castings. We do have some ideas here, however, and we will be trying them on the forthcoming battleships (apart from the Mars-class, which is all metal, just like the Federation dreadnought). Success here will mean we can go back to the dreadnoughts and redesign them.