What are you trying to playtest, in particular?
Start with knowing what you are looking to test, in this particular playtest run. Obviously, other things may come up, but know what you're looking for as the primary thing to be tested - core mechanics, strategies, polishing edges of rules, flow, etc.
You may want to tell the group what to focus on, as well, since that will help them be more mindful during play, and take notes on that.
You mention having only the rules sent to the group as the form of contact. This would either be late in the playtest setup or with people you know. Early on, you want to test core rules, etc. and you don't want to spend a lot of time dealing with miscommunication or pedantry.
Fun and Flow
First, I look for groups to report what was fun, and how things flowed. This isn't a particularly deep or specific kind of thing, but you're looking for general signs to see if their experiences are matching up to your goals.
It gives you a bit of an angle to how you read/listen to their playtest report - if it matches up, then you're looking at technical solutions and polishing of bits. If it doesn't match up, you either have serious design problem or serious problem in communication about the design - so you have to step back from looking at details of mechanical assessment and more towards a broader picture.
What rules were unclear? What parts did the group get lost as to "what to do next"? What felt hard to run? Did anything seem unfun?
These again may be design solutions, or communication solutions. This is also where you end up throwing a good amount of feedback out. "Your game about tactical fighting combat didn't have enough rules for romance!!!" and other such things come up, where the only thing you can do is make sure you're communicating the goal of your game well. (Vincent Baker talks about this as the Object of the game.)
The way to pick apart specific mechanics issues is to follow this line of thought:
1) "The thing I wanted to do in the game is X. I tried to do Y to make that happen."
2) "The mechanics to do Y make it (easy, hard, a tough choice, are too complicated, no choice at all, etc.) as a process, and those mechanics make me want to do Z."
Your questions/discussions with your playtesters basically looks to pick this apart. Understand that, unfortunately a good number of gamers are not particularly adept at this.
"I wanted to have a strong fighter-character. I tried to take this set of Special Sword Skills to make that happen. The mechanics to use Special Sword Skills use up my Hit Points, which makes me actually not want to get into fights after using it."
Once you have that, you need to think for yourself if Goal X fits with what your game is supposed to be about in the first place, if Method Y even fits with Goal X, and if result Z is also in line with any of the aforementioned concepts.
Sometimes this means better communication, sometimes you just have gamers who want your game to be something else, and you're not going to get much value from that communication.