If I were looking at something like this I would rather use "Most recognized" metric on bpip, it at least evaluates more factors and not just merits.
This factor is heavily in use to separate doubts between two or more equal choices and pick one from them.
Not to mention there are other factors like
- Having people from busy local boards and balancing how many a campaign can afford to have from a same local board
- Users have habit posting in specific boards. We don't want a poster who have all concentration in only one board.
- Users knowledge about Bitcoin is a good factor too. We want knowledgeable bitcointalk members in a campaign.
I hope that we won't have huge oscillations in payments with merits as the sole evaluating criteria.
Merit definitely helps to filter between good and bad posters but when searching for THE BEST people, we need to find the good posters first. The Merit rank helps a lot in this case.
Regarding actual question and amount of posts that will be paid I don't have strong opinion. I am now paid for 25 and I am writing about 35 while being very busy in real life. That will probably rise to my standard 45 when workload drops to normal no matter how many posts I am paid for.
In my opinion, being in a campaign does not mean a member limit himself to the highest paid post per week rather he needs freedom and have the opportunity to get paid for all the GOOD posts he is making (without a ceiling of maximum posts to get paid) to contribute in the community.
To prevent the possibility of spamming, having a maximum count of 40 would be better than having 50.
And thus, naturally, we first go with 200$ for 40 posts, and if we can control the spam possibility, we raise the bar to 50 posts and the limit will raise to 250$? [this is assuming that I assume correctly the new payment structure you'll use]
I like the idea to experiment how $200 per week goes then try $250 per week. But I can assure you that when a campaign have a budget to pay $200 or $250 per week, spam / bad posters are the last thing that will be tolerated.