Paleoanthropology is a tricky field. When all you've got are a specimen or two from one site (or worse, just a partial jawbone or two), you try to get as much out of it as you can. But until more specimens turn up for comparison, it takes a while to figure out if what you're seeing falls within a range of natural variation, and eventually you might be able to say, "Hey, this isn't just variation within a species, this is a different species." If you're really lucky, you might find a whole bunch of individuals at the same site, which helps to compare the variations between males, females, adults, and juveniles.
But agreement in paleoanthropology is slow. I mean, really slow. People can be disagree about where specific fossils fit into the family tree for 15-20 years. So these early announcements should be taken with a huge grain of salt - but it doesn't make them worthless. Paleoanthropology has a lot of very creative thinkers who bring as many ideas as they can to the table, which at first makes it a bit of a muddle, but as time goes on and more fossils are found, you whittle away those ideas, rejecting some, tentatively accepting others, keeping a few in the "Mmwell, maaaybe..." pile.
There are also long-term trends in the field. The early 1900-1940 paleoanthropologists accepted evolution but many didn't accept the mechanism of natural selection, had racist interpretations of anatomy and disregarded Africa as a good place to go looking for ancestors. Variation within species was often ignored, so any little difference in a lump of skull resulted in a new species being announced - "Splitting" the family tree. Sadly, most of the fossils they were looking at weren't older than Homo Erectus, with an Australopithecus or two thrown in, but being disregarded.
The next generation of paleoanthropologists, from about 1940-1980, benefitted from the discoveries of what's now called the "Modern evolutionary synthesis". A better understanding of natural selection and variation was back, racism was frowned upon (although there were some hangers-on), Africa became a huge focus, and a lot more Australopithecine fossils were finally giving the field some much-needed depth. "Splitting" was out, "Lumping" was in. The earlier generation's work was largely considered bunk, and all the proposed new species were lumped into a small handful. This gradually became a problem for new fossils; the lumping paradigm was so strong, it was very hard to propose new evolutionary branches. It didn't help that Louis Leakey stayed in the older tradition and kept naming new fossils in whatever way he liked. (It's also worth mentioning that his wife, Mary, was the person behind a lot of discoveries, but Louis was the one who mugged for the journalists.)
Then during the 1970s and 1980s the next generation started to have an effect - a lot more creative ideas began to emerge. Theories about behaviour, tool use, looking beyond anatomy; feminist challenges to hunter-gatherer models; Gould's punctuated equilibrium ideas; Wolpoff's multiregional hypothesis. Unfortunately there was also incredibly poisonous in-fighting thanks to Donald Johanson, Tim White and Richard Leakey, which trickled down to their students and made for terrible politics and competition for African sites, which was still being felt thirty years later. "Splitting" gradually came back in during the 1980s; the field is still in a Splitting phase right now.
And to really give you a scope of how long it takes paleoanthropologists time to come to agreement, the position of the Neanderthals is still under debate. But such is science! It'll take as long as it needs.