Nick said:
But the idea of a large surface area is a bad one...
1) Doubling the surface area double the deceleration - the "g" loading. Existing capsules pull 3 g. your saucer has about 100x the base area to volume ratio of a compact capsule - leading to 100g re-entry loads.

I would agree if the decent trajectory remains the same, but let's suppose that the decent slope is 100 times shallower, then the "g" loading will remain the same. Naturally the heat shield will need to take care of the extra time involved in the de-acceleration.

If your capsule was strong enough to take this load (strong =heavy, BTW), THEN it would flutter gently to the groumnd and you could scrape out the contents and try again. :-(
If the decent slope was shallow and the heat shield was able to withstand the duration of the stress, then the strength and weight would not need to increase.

2) wide flat shapes are intrinsically unstable. You DON'T want to flutter during reentry since this leads to violent shaking, irregular heating, and possible destructive forces requiring a strong and heavy airframe again, and more astronaut jam.

Wide flat shapes are intrinsically unstable unless they are spinning, but cone shapes are quite stable especially if 'supported' from the sharp point.

Keep smiling
Johannes