"I've complained before about the undeserving admiration given to many science fiction novels and short stories of earlier decades. Such reverence frequently results in fans recommending these works to people wanting to try the genre. And that's not a good thing. Readers new to the genre are not served well by recommendations to read Isaac Asimov, EE 'Doc' Smith, Robert Heinlein, or the like. Such fiction is no longer relevant, is often written with sensibilities offensive to modern readers, usually has painfully bad prose, and is mostly hard to find because it's out of print. A better recommendation would be a current author - such as Richard Morgan, Alastair Reynolds, Iain M Banks, Ken MacLeod, Stephen Baxter, and so on.
I can hear howls of outrage across the tinterweb."
There's a wider argument behind what Ian has to say. For people outside of the genre, authors like Clarke and Asimov are familiar touchstones they've heard of regardless of whether or not they've actually encountered their fiction.
However, these writers also date from a period when many writers - and Asimov was particularly vocal in this context - who believed that the Idea was far more important than the language used to express it; in other words, the 'story' was nothing more than a vehicle for the Idea. If the prose was a bit rough, well, that didn't matter as much as the Idea.
This is why sf has historically been seen as a low-rent form of writing typified by bad prose skills and wooden characterisation. The fact that much of it was written in a time when open racism and deeply offensive views of women were much more common currency can at times become unpleasantly obvious. Elements of this can be found, unfortunately, in some of Heinlein's novels.
This changed to a great degree first with the 'New Wave' sf of the Sixties and later with publications such as Interzone. The quality of the prose became as important as the Idea, as did developing rounded, more interesting and - particularly in the wave of New Space Opera - morally ambiguous characters. In all, things have become much better, and it's rare for me to find pleasure any more in the old-school fiction. I find that I overwhelmingly prefer to read stuff dating from the late Seventies on, with notable exceptions such as Ellison, Tiptree, Dick, Delany, Moorcock and others of similar ilk.
On the other hand, the reason some prefer the old stuff is quite simply because it's relatively straightforward, dating as it does from a time when the available scientific knowledge was still relatively easy to assimilate in a diluted fictional format.
Some critics, such as Barry Malzberg, have argued that the average non-genre reader might well have no idea what is going on in a modern sf story if they were to pick up a magazine and attempt to read the contents; that the concepts and language used to express the Idea can be particularly opaque to a casual browser with relatively little conception of such notions as quantum physics, nanotechnology or dark energy.
The answer for new readers, I think, lies in a middle-ground: accessible, well-researched and well-written fiction. Kim Stanley Robinson immediately springs to mind in this context. So does, despite the density of the ideas involved, Neal Stephenson. Ian's suggestions of Richard Morgan and Iain Banks are apt ones, in my opinion, as would be Dan Simmons and also early to mid-period John Varley.
This all isn't to say that all the older stuff is rubbish, far from it; perhaps the more accurate argument is that the older fiction being recommended is the wrong kind to recommend. Rather than the Damned Holy Trinity of Heinlein, Asimov and Clarke, we should be recommending Silverberg's Book of Skulls, Zelazny's Damnation Alley, and Philip K. Dick. Perhaps an apt metaphor for the difference between the two forms would be: Old Testament SF, and New Testament SF. The old, mean, lightning-bolt chucking stuff as opposed to the later, more ambiguous works.