SNPs Inside STRs?

On September 4th, 2014, Roberta Estes posted a blog "What Does and Doesn’t a Y DNA Match Mean?."  It was chock full of useful information helping answer some of the most frequently asked questions of beginners in yDNA testing as well as those who've been at it a year or two.

Warren Power of Australia, admin of the Power et al DNA & History Project, gives a great explanation of what STRs and SNPs are.  He quotes or paraphrases from a 52-page booklet by Kerry Farmer.  I've given an excerpt from Warren's email which you can read in my TEXT file SNPs vs STRs.

One question that usually goes unanswered is aluded to by Ken Nordtvedt's Aug. 2014 spreadsheet on I1 Modalities.  You don't have to understand the spreadsheet's math to get the point I'm trying to make.  SNPs are on the left in the 1st column, STRs are on the right in rows.  There's a Col. vs Row relationship here.  That means there's a one-to-one relationship between SNPs and STRs, right?

Ok, maybe I don't know what I'm talking about but I know at least some of what I'm saying is correct.  So the first question that crosses my mind is, "What exactly IS the relationship between SNPs and STRs?"  I know they're two different things.  I am NOT suggesting otherwise.  They may likely exist in two entirely different world's for all I know.

I only wanted to know, "Is it possible for them to exist in the same world?"  More specifically, "Can a SNP exist INSIDE an STR?"  Simple question, right?  WRONG!!!  It seems no one has ever asked that question before.  If they have, I wasn't able to find the question or its answer on the net anywhere.  Believe me, I searched.  So I asked Roberta.  
As they say in the funny papers, "That's when the fight started."  Just kiddin'.  No, there was no fight.  And Roberta is quite the expert.  But she didn't know either.  So she asked Dr. David Mittelman who has been the Chief Scientific Officer of Gene By Gene, Family Tree DNA's testing company, since July 2013.

I was specific asking Roberta this exact question.  Roberta is one of the most honest bloggers out there.  If she doesn't know the answer to a question she says so.  Then she finds out.  She would let us know when she got Dr. Mittelman's response.  She did and Posted It On Her Blog in the Comments:

A SNP can certainly exist in a repeat tract. You could have something like this:

CAGCAGCAGCAG (4 CAGs)

CAGCAGCACCAG (4 CAGs, but the third triplet has G mutated to C)

I would still describe both repeats as CAG(4). It is true that the first one is a perfect repeat, and the second is not 100% pure repetitions. I believe at the nucleotide level this happens for at least some STRs, they are not always perfect repetitions.

NOTE:  Comments are in chronological order.  My question was asked September 9, 2014.  Roberta posted Dr. Mittelman's reply on September 11th.

WOW!  That not only explains a lot about the relationship between SNPs and STRs, it also tells us much about how genetic researchers view STRs to begin with.  The STR repeats, what we usually refer to as "Markers", may not be as pure as we had previously thought.  

Dr. Mittelman suggests that individual nucleotides within a particular marker may differ slightly from its fellows and still be considered as a repeat.  If not, it could change one's STR numbers, couldn't it?  The next question to ask is do all researchers accept markers that "
are not always perfect repetitions
"?  But then that's why we always recommend sticking with an accepted lab like FTDNA rather than going with some of these other DNA testing companies, isn't it?