The problem is that you can't compare a PPP poll with a Monmouth poll, for example, and use the difference between the two to support a conclusion either way. They have different polling methods, different sample sizes and demographics, different analytical methods, etc. You are virtually comparing apples and oranges.
If you want to see how movement is trending, you can either compare different polls by the same companies (as I have done), or you can look a the polling averages of an aggregation of different polls, which helps show a trend without worrying about the specifics of any one poll. Using either methodology, the recent polling shows a movement toward Carson. That may change, but misrepresenting the results of these polls does little to enhance the credibility of those doing so.
So now we have two with Trump in the lead. I want to see the same orgiastic response. Waiting . . .
Both of those polls were already showing Trump in the lead - one is showing a slightly larger lead, the other is showing a significant decrease in his lead. But neither of them is showing a significant increase in support for Trump or loss of support for Carson...
But mo seriously, yes, you can cross reference polls IF you can figure out which ones are "leading" and which are "lagging." I don't know which is which---but the likelihood, given all the polls, is that the Carson boomlet is over.