In the 30 years since Viking many items have decreased in cost or alternatively delivered much more bang for the buck. NASA has made minimal gains over that period. Contrast the scope of this 2008 mission with Viking's. Never mind that the V mission planners did not have any prior Mars orbit/landing experience to draw from.

While electronics have made radical improvements bang for the buck in the past 30 years, not many other things have. There are better materials available to build spacecraft but most of them are very expensive. It has been claimed that the most advanced rocket engines are made in Russia, largely because NASA was directed by several presidents and Congress to go with the Shuttle. After the SSMEs and related engines were developed, NASA really didn't do much to spur new engine development. Expendable boosters are still damned expensive largely because the manufacturers have little incentive to lower costs.

When it comes to spacecraft, some components have improved radically in the past 30 years. Solar arrays are much more efficient now and heavier lift capability means that much bigger satellites can be launched. For example, 30 years ago (1973), the biggest geosynch satellites probably weighed less than 1,000 pounds. Today, the biggest ones like Milstar come in at over 5 tons. Sensor technology is a lot better now than it was 30 years ago, but not necessarily less expensive. It still takes a lot of R&D to develop new sensors and good geeks don't come cheap. Better pointing accuracy and improved communications technology has allowed for higher data rates than ever before, but it's still tough to get much more than 10 kilobit/sec of data at interplanetary distances. A lot has changed in 30 years but fundamental physics isn't one of them.

Contrast ISS bang per buck with SkyLab's.

Let's not forget what Skylab was. Skylab was largely built from surplus Apollo hardware. It was launched on the last Saturn V mission and damn near failed when a solar panel and heat shield were ripped off during launch. It didn't have any propulsion system and wasn't designed to be refurbished, so it's operational life was limited to a less than a year. It's an apples and oranges comparison to compare Skylab to the ISS (and remember, I'm no fan of the ISS).

Where NASA has continuously improved is with the science yield from their astronomical observatories.

I'm a big fan of the orbital observatories but only for those bandwidths that don't penetrate the atmosphere. New terrestrial observatory technologies like adaptive optics allows for optical resolutions and sensitivities that rival the Hubble and they cost a fraction as much. For the cost we've paid for the Hubble (over $3 billion according to some sources), we could've built at least a dozen cutting edge terrestrial telescopes. Which would give the best bang for the buck? Of course, for those wavelengths that don't penetrate the atmosphere very well, there's no substitute for orbital observatories. For those, I'd like to see a complementary set of ground and space based observatories to cover RF and IR through gamma so that when events like supernovas happens, we can gather data seemlessly across the entire EM spectrum.