The simple answer is that the fact that Marco Polo went to China doesn't preclude him from lying about or distorting what he did there. The more complex answer has to do with the ways that medieval texts were written and replicated.
The problem of the siege of Xiangyang has been used as a major 'gotcha' for the radical anti-Polo position for some years, and on the surface it seems pretty damning: if Marco Polo went to China in 1275, how was he involved at a siege in 1273? This was an event where although foreign expertise was called upon for building siege engines, according to the Chinese sources it was Muslim engineers, not Venetian merchants, who provided that expertise. Given that the position of anti-Poloists like Frances Wood has hinged mainly on the idea that Polo merely adapted now-lost Persian geographies and chronicles, it would seem to be a pretty obvious mark in favour of that position. However, there are a few arguments to be made to suggest that the Xiangyang incident is less damning for Polo's account than Wood and others have made out.
Peter Jackson (no, not that one), in his article on the Book of Wonders, notes that Polo consistently exaggerates his importance within the Great Yuan, for reasons either cynical (in looking to inflate his own importance to the reader) or naïve (he may have interpreted being sent on minor errands as being of greater prestige than his Mongol handlers did). Polo could easily have made the claim up himself, inflating what may have been some sort of involvement in a lesser military affair into direct provision of expertise at a major siege.
But another thing Jackson points out, with reference to issues that had already been flagged up as early as 1958, has to do with the problem of the manuscript tradition. With any medieval text, it is always worth noting that there is rarely ever a definitive edition of anything. Texts were frequently copied, and modified between copies; some texts might also be revised by their authors over time. Arguments over which manuscripts came first are not easy to settle. Ronald Latham, translator of the Penguin edition of Polo's Book of Wonders (whom Jackson cites), notes that while the detail about the siege is present in a number of the manuscripts, it is absent from others. While the manuscript considered the most reliable, the French manuscript "F", does include the detail, it is absent from the Venetian "V" manuscript and the Latin "L" and "Z". This is significant because while "Z" likely postdates "F" and is likely an abridged form of some text, this abridgement seems to have been derived from a fuller Latin version that predates the composition of "F". While an absence from "Z" does not prove an absence from the earlier version, that the Latin "L" manuscript (which is often used to supplement "F") also elides the Xiangyang case suggests that the detail about Polo being at Xiangyang came about in one particular strand of later versions of the text, and was not original to Polo's own narration.
Hans Ulrich Vogel, in his 2013 book on Marco Polo, does not mention the manuscript problem, but concurs with the critiques of Wood by Jackson, de Rachewiltz, and Morgan in noting that Polo was fully capable of lying. However, Vogel includes in a footnote a summary of an argument in a Chinese book by Peng Hai from 2010, that Polo could have been involved at Xiangyang, again with reference to the manuscript problem. Peng's argument includes the following elements:
Vogel notes that he does not find Peng's argument convincing, but leaves the argument against for another potential publication.
In short, then, Polo's supposed involvement at the siege of Xiangyang need not be an inconsistency that severely detracts from his overall reliability. For one, Polo could easily exaggerate some parts of his narrative without fundamentally compromising the whole. For another, the simple fact is that the manuscript tradition does not unambiguously place Marco Polo at Xiangyang in 1273. If you don't necessarily buy that Polo was lying, you could go by Peng's (somewhat more optimistic) argument that it is plausible Maffeo and Niccolò Polo went to a later siege that became conflated with the siege of Xiangyang in some uncorrected manuscripts. More realistic, though, is Latham's suggestion that the detail of his involvement at Xiangyang derives from later additions to the manuscript tradition that do not come from Polo himself.