So this model has been attempted before in various formats. As a direct comparison, 3-4 years back there was a company called WidgetBox. They were a startup. Successful in getting funding. Raised at least $8 million. Their changed up their model a few times but their most successful one was nearly identical to what you described. They went directly to various advertisers on a CPA basis and then guaranteed publishers a set CPM based on the agreed CPA with the merchants. Got as high as doing 500 million impressions a month. But they didn't appropriately account for fraud, had to back out on payouts, ended up nearly folding. They were able to pivot and be absorbed into Flite.

A less direct comparison of your scenario is very common. Many affiliates these days operate what is considered a sub-network (against the rules of most larger affiliate networks) or a super-affiliate program. Examples are the dozens of loyalty affiliates out there like Upromise who also have their own affiliates (as well as members tracked on sub-ids) underneath them.

Being the advertiser's "sole" affiliate is partially where I don't see the model you describe work. Unless your advertisers are completely unfamiliar with the digital space they are unlikely to only work with one company as their sole affiliates. Advertisers like to scale. It's why they work with networks.

What ever you decide, Post Affiliate Pro does not have a robust enough of a platform for you to launch with. Beyond that the software's ability to help detect fraud is suspect. HasOffers (know called Tune) is a way better choice. Also recommend looking at Performance Horizon Group.

Either way, highly recommend rethinking the "exclusivity" or "sole" component of your model and asking yourself why an advertiser would just go with you?