This paper is a good example of jumping to far reaching (and perhaps erroneous) conclusions from very limited data. Several things need to be noted from the news items (I can't access the original paper at this stage).

1) We don't know the age of the olivine-bearing material at Nili Fossae. It may not be 3 billion years old. In fact, Nili Fossae is regarded as Amazonian, the youngest epoch of Martian history, less than 1.3 Ga

2) The area is 30,000 square km, equvialent to an area about 200 X 150 km, a moderately extensive basaltic lava field. Extrapolating to the whole surface of Mars, which consists of many different types of materials (sediments of many types, volcancic flows, intrusive rocks, primordial crustal material) in many different settings, altitudes and latitudes is unawarrented.

3) Interpreting mineral composition from remote spectroscopy is very tentative and the reuslts can be misleading. TES has never been calibrated in the field, from either ground or airborne sensors.

4) The news items are in error when they say that weathering products have not been detected on Mars. Must compositions from both the Viking sites and the Mars Pathfinder sites are consistent with highly weathered materials, as are the iron contents in the GRS Mars Odyessy data.

5) The news items exaggerate the instability of olivine. While it does weather readily, it won't disappear whn you drop it into water. Olivine is used in water-based abrasives and is found in beach and river sands, sometimes in high percentages, having survived for thousands of years in warm, wet, saline environments. It can be found as a detrital mineral in ancient sediments. Olivine can also survive in the the terrestrial surface environment for millions of years in igenous rocks.

Jon