Except the ideal one doesn't exist and multistage stream turbines are more efficient in practice.

Definitely not at that scale. Small turbines have low efficiency because tip speed increases quadratically as they get smaller.

And more reliable too. The Stirling does have one huge advantage which is that it is entirely self contained.

But I suppose that having a multistage steam turbine in mars in not something will see at first

You may be interested in reading Allan Organ’s books. He claims that aspirations to Carnot efficiency are a red herring and claims (with Much Math) that the peak efficiency is closer to 50% of The Carnot cycle.