The SR-71 was an impressive engineering feat because it successfully brought together a bunch of really crazy tech and got it to work. All of this with a relatively small team and stone-age (by modern standards) computational ability.
The A-10 was an impressive engineering feat because it perfectly executed a very boring and conservative design. It did one thing, did it very well, and was so reliable it arguably put its producer out of business.
If I had to draw a comparison to software, I'd compare the SR71 with Erlang, and the A-10 with Postgres.
They were both such impressive pieces of engineering execution that they completely changed their respective games. They may be old now, but they are engineering success stories worth admiring.
