The Ghost from The Grand Banks is one of the last of Clarke's books to be published before he began prostituting his name with a stream of substandard "co-authorship" efforts. For long-standing Clarke fans, it's the last of the good stuff -- but not by any means the best of the good stuff.
For good or bad, Ghost epitomises Clarke's classic style: creaky, one-dimensional and distinctly old-fashioned protagonists wander through an idealised world of scientific and technological marvels, only occasionally troubled by brief and frankly somewhat embarrassing flurries of characterisation and motivation.
Many of the character tropes are over-familiar to the fan: a world-weary engineer/adventurer, the partners in a cold marriage, Oxbridgesque Johnny Foreigners and a gentlemanly toff. Many of the characters are interchangeable and disposable, even more than is usual in a Clarke novel: at times it feels like none of them at all serve any function other than to provide wryly-delivered exposition.
This much will not be too depressing to long-standing Clarkeniks; we've long since inured ourselves to this state of affairs. What is more irritating is that this time round, the scientific marvels tend to be a let-down. Ghost has not dated well, perhaps because it is set much closer to the present day (2012, the centennial of the Titanic's sinking) than many of Clarke's other Great Engineering novels -- and its time is ten years closer now than when the book was written. Some of Clarke's predictions and factoids have already come true, and are thus boring; the others either turned out to be wrong or are no longer interesting. At the time, perhaps it was reasonable to think that a lengthy and plot-irrelevant diversion through the Mandelbrot set would hold readers' attention. Today, it isn't.
One also has to question if the engineering-heavy style is really the best way to proceed with this story. The appeal of the Titanic story as the inspiration for fiction is surely a human one. Judging by his other work, Clarke would be hard-pressed make it sing even when on form; and here, he isn't. More than anything else, the book suffers from a lack of integration: it's a catch-all collection of nifty ideas that might have been better separated out into novellas (if only Clarke, he of the 850-word "novel outline", were to do those any more) and given some narrative drive.
To sum up, this is a poor excuse for a novel. It's a thinly-fictionalised description of some schemes for raising the Titanic, bundled with some more-or-less disposable extra stuff that was moderately nifty ten years ago. Clarke nuts (like this reviewer) will find it a moderately diverting quick read: three stars. Others might find The Fountains of Paradise or Imperial Earth more worthwhile: one star. So on average, two stars.