I'm no expert but this is what I learn from reading and my own research.
PoF turn out to be OK for me.
I'm in breakeven after spending almost 200k impression for a campaign.
Got $64 for $60 spent on testing 15 different creative and targeting different demo.
(testing and breakeven = collect free testing data)
Optimizing and weeding out ads with low CTR and splitting the demo to optimize traffic after 500k impression
IMO, PoF traffic converts. You just need a good target with a good LP.
Try targeting traffic from different angle. I think your AM can give more advice on that tbh.
As for impression, I cut my ad on different stage
1. 5000 impression - any ad less than 0.02% CTR
2. 10000 impression - any ad less than 0.03% CTR
3. 20000 impression - any ad less than 0.05% CTR
My ad cutting method is very loose because I use high payout offer, a few of my ad has 0.2% CTR and my EPC is $1+ to $2+.
For dating offer, that might be very different as the dating offer EPC might go down as low as $0.35.
Some mentor suggest cutting ad that has CTR lower than 0.1%. That might be problematic if you are not an expert in making ad copy.
One thing to avoid is cutting ads too early.
(CTR can change quite drastic when you have a small number of impression e.g. 1 click at 5k impression change CTR by 0.02%)
While testing I use one or two different text (initially one and add another one later in testing)
I use different image to gauge CTR attraction. Then I'll make changes to "Ad Title" and ad body.
As CTR at 20-30k impression is quite stable, you can see the effect of changing ad title and ad body significantly.
Hope that helps

and don't forget to track your ad copy