So back in the summer of 1935, some folks down in Australia were having problems with a beetle's larvae that were nomming on the sugar cane roots and harming crops. Since traditional methods of getting rid of the pest failed, they decided a good approach would be to introduce a few cane toads to go eat the beetles... a hundred or so of them in a couple of places in the north east. Seemed like a good solution.
Fast forward to today, there's now hundreds of millions of the things happily munching their way across the country, even evolving traits that help them move faster. They are spreading at a rate of 60km/year, trashing the ecosystem of everything but the dratted beetles they were introduced to combat. See the problem is that they're toxic and none of the native biosphere had evolved in the presence of that toxin, thus they had no natural predators and any potential predator deciding to eat one of these toads is likely to eat little else afterwards.
What does this have to do with space? Two things.
Firstly, this is an example of an introduced invasive species... in the case of the toads it was intentional, rabbits are another species that has had similar disastrous results in Australia's ecosystem. There are also examples of unintentionally introduced species, with much the same result. If space probes successfully transferred viable micro-organisms to another planet we have the risk that any micro-organisms already existing there will be overrun by our earth bred critters. That would be bad.
If we can mess up a relatively similar ecosystem with something as large and predictable as a rabbit, how much of a mess could we cause by introducing microscopic and unpredictable lifeforms to an ecosystem we know nothing about?
Secondly, the purpose of a lot of our planetary probes is to locate signs of life. Ideally currently live life, but signs that life once existed will also do in a pinch. This is kind of difficult if we're sending probes that are covered in lifeforms. It's equivalent to making a radiation detector, putting it in a uranium housing, and then releasing a study on the elevated background radiation in your location. You can't reliably detect something if your detection equipment is contaminated with the very thing you're trying to detect.
And even if the probe that is doing the detecting is clean, if another probe isn't then the life you detected might have piggybacked on that. If you want to say for certain "this life evolved on Mars" you need to ensure that it definitely didn't arrive on a probe.
Worst case, you introduce some lifeform to the planet and it promptly starts eradicating the life you're looking for. Red squirrels are the native species in the UK but with the way the introduced grey squirrel has been pushing them out for decades, if you were looking for signs of squirrels here then it's possible you might never encounter evidence of red squirrels.
The bottom line is that it's unlikely that such a contamination could happen but not nearly as unlikely as you'd think. If you send millions and millions of viable lifeforms or spores in to space, some of them are going to be able to survive, some of them are going to be in a spot that the solar radiation isn't hitting, some of them are going to be inside a shielded bit, some of them ... you get the idea. Accidental sterilisation isn't in the slightest bit reliable enough and the potential problems posed by the two reasons above are sufficient to justify the effort spent on trying to reduce the risk.
The argument that life could travel from Earth to Mars via meteorite is valid, but actually demonstrates why planetary protection is so important: If you find Earth micro-organisms on Mars, how do you know if they got there via meteorite and not via a probe? Knowing if life could make that trip would be very useful, in theories such as those involving the inverse, life being introduced to Earth via meteorite.