Hi,
I think you're looking at these spots from a very narrow view point of make-a-hand poker (stage 1 thinking).
It's commendable that you want to play a disciplined game, but in heads up holdem, this won't likely be very profitable.
Why? Because over the long term, you and your opponent statistically get dealt the EXACT same distribution of hands. Your edge doesn't come from making a hand and betting it - both of you will do this with your made hands. Your edge comes from finding spots where your opponent has particularly exploitable tendencies and taking advantage of them.
For example, if I'm playing you and you're a guy who folds A2o @ 30BB to a min raise => I'm going to take advantage of your frequencies by raising 100% of hands from the button. You would be letting me make money from hands like 32o. Sounds ridiculous doesn't it? Don't be that guy.
You need to think beyond: what is my hand, what sorts of hands does it make?
In fact this is probably less important than your position and your opponent's tendencies with respect to maximising your expectation.
 
1. Don't fold A2o @ 30BB.
The only type of player you might entertain this thought against is someone who only opens an extremely tight range of hands.
The majority of players open enough that this is a decision between 3b (leveraging your blocker to his raise-call range) and calling.
First consider these questions:
a) What is my opponent's opening range @ 30BB?
b) What range would he continue with against a 3b?
c) Is he capable of 4b jamming light?
d) What is my image? Perceived 3b range? Perceived calling range?
e) Gameflow? How active have I been?
f) How often does he cbet with his air?
g) How aggressive is he postflop? (will he barrel his stack off on A high boards with his air?)
 
^^ If you're not thinking about these tendencies, you're leaving a lot of expectation on the table because these will determine your best course of action and your plan for the rest of the hand.
 
So do we flat or 3b?
Some general guidelines with respect to the above:
a) the wider he opens the more equity your A2o has and the more likely he will fold to a 3b.
b) the tighter he continues vs 3b, the more likely he will fold to a 3b.
c) this is a balance between whether villain is spazzy enough that we can 3b-call (usually not) or 3b-fold with A2o. Typically calling would be better vs these aggressive opponents (explained below).
d) if your opponent discounts any Ax in your flatting range then flatting can provide a lot of implied postflop value. If you are perceived as tight then your fold equity should be better if you 3b. Your perceived range will also allow you to attack board textures which "should have" hit your range. IT DOESN'T MATTER THAT THEY DIDN'T BECAUSE HE DOESN'T KNOW THAT :)
e) villain may adjust his ranges depending on how he perceives how active you've been.
f) the more villain cbets relative to his opening range, the higher his bluff/value ratio in his cbetting range and the profitability of c/r bluffing goes up.
g) the more aggressive the opponent is, the more he will donate implied value when you hit your A. If villain has a tendency of firing a cbet then giving up without the goods, A high has fantastic equity on dry boards => you can profitably c/c then check down to exploit his passivity.
 
2. 65s @ 21BB
As above, you need to build your decision making framework beyond stage 1 thinking.
It's probably not going to be a large loss in expectation to fold this hand OOP facing a min-raise @ 21BB but if you can find a way to play the hand for EV > -1.0BB/h you yield from folding vs your opponent, then you should take it.
 
FYI: Readless I think flatting A2o @ 30BB and 3b NAI 65s @ 21BB will have the best expectation but your decision should be framed within opponent-specific observations.