I'll try to summarize why the evidence, in this case and other similar ones, is considered not conclusive.
The scientific community agrees on America being populated as late as 13-17k years ago, and as early as 40k:
The time range of 40,000–16,500 years ago is a hot topic of debate and
probably will remain so for years to come. The few agreements achieved
to date include:
- the origin from Central Asia
- widespread habitation of the Americas during the end of the last glacial period, or more specifically what is known as the late glacial
maximum, around 16,000–13,000 years before present (One of many sources: Did First Americans Arrive By Land and Sea?)
This is why those affirming Hueyatlaco is older than that (link to most comprehensive publication about the subject) are generally thought to be dealing with contaminated evidence. The biggest defender of this theory is Virginia Steen-McIntyre, who was a graduate in the 1966 excavations.
The way these sites' antiquity is measured is usually 'indirect', meaning that if stone tools are found in a certain stratum, it's the stratum that gets dated, not the tools. The stratum where stone tools were found in Hueyatlaco was dated as older than 35,000 ybp by radiocarbon (animal or plant remains - only good for 'newer' datings) and 260,000 ybp, ± 60,000 years by uranium dating.
The discussion around the dates has to do with how trustworthy an indirect measurement like this is, when the value falls so off what's generally agreed on. You can't date the tool itself because stones don't give any information about time of human manipulation (unless you have a known classification of technologies or styles).
The problem is that strata is not necessarily something static. Steambeds and other natural phenomena, as well as animals digging caves, for example, can erode and change their structure. There is a number of documented reasons why you can end up with newer artefacts on older layers (I worked on a cave site that was almost impossible to date this way, because there were rat skeletons in every stratum, showing these animals had dug caves and moved all the materials around). The area around Hueyatlaco is subject to flooding.
So, unless you have tools can be dated by themselves (made of organic materials, for example), there is always space for speculation of contamination. The rocks in that strata are that old, yes, but the tools on it seemingly are not. Affirming they are would be debunking what we know about Homo sapiens.
Because the current paradigm states that the first populations for that part of the continent moved in about 13k to 40k years ago, these datings are considered to be not trustworthy.
Some more references regarding archaeological contamination and the relationship between archaeology and geology: