Purely hypothetical...but it seems to me that if one runs in certain circles, one would know plenty of people willing to pay 90, 80, or even 70 cents on the dollar to clean/hide their money. Maybe I'm missing something, but a steady stream of dirty funds wanting to be cleaned would certainly allow pirate someone to pay his 7%/week by running a BTC laundromat - no ponzi necessary. Further, I think running a laundering scheme via BTC would be pretty durn complex...i.e., it couldn't be unwound quickly/easily/without significant loss if something unexpected occurs.
Others have already debunked this business plan based on the plain unnecessary costs of choosing to lend money at 3400%/year interest when something like 10%/year interest is much better choise.
However that is not the only argument to debunk this business plan.
The problem is that maybe you looked at the 30% profit of selling the coins and thought that when it is bigger number than the 7%/week interest, then the business is making money. Well it is not. The 30% profit is a one-time profit of selling the coins, but the 7%/week interest keeps on running till infinity until the debt is cleared. The coins that pirate has sold still remain as coins that pirate has lended and he needs to pay 7% for them on this week, another 7% the next week and 7% the week after that and pretty soon the whole 30% one-time profit is eaten and pirate has massive problems keeping paying the interest with no new profit in sight with those coins he has long sold. If you think that pirate solves the problem with new customers for more coins and pirate then lends new money to sell them that means running the traditional "running a massive loss but making it up in the volume" business plan.
I cannot think any other solution to the problem but pirate needs to start rotating his entire capital. Not just new money, but everything also from the beginning. Use the obtained cash to buy massive amounts of coins and sell them at profit again. This might work if pirate would rotate his entire lended capital every 2 weeks raking in new profits every 2 weeks.
However there are big problems in this sceme as well. The primary motivation to all this is money laundering, right? Typically money laundering works in an trivially normal established business which does not raise any eyebrows if somebody is caught involved in that. Additionally the volume of the global business must be so huge that the laundered money does not blip on the radar.
Now imagine that pirate needs to rotate his entire capital of 500k BTC ( or whatever ) in every 2 weeks. Annually it means that pirate would need to buy at annual level about 12500k BTC and sell them to the buyer which needs then to sell them in some place so that the fiat money coming from the end looks like it is plain normal vanilla money with nothing suspicious ( Where could he do that? ). 12500k BTC means USD125M money flow for pirate to buy coins with and then somebody to cash out. Sounds slightly nontrivial and if IRS gets hold of the wind, mr. pirate might be in trouble explaining where is he getting money to buy USD125M worth of bitcoin every year.
Somebody could think this whole business plan sounds ludicrous.
The exactly same argument can be said about any other business plan which involves a one-time profit. The massive profits need to rake in every couple of weeks. Using the lended capital to controlling the BTC price is another business plan where I don't see the frequency of the massive profits to be high enough and also needs to rotate huge amounts of capital very fast to obtain enough profits in this scheme as in here the profits can not be expected to be 30% per every rotation of the whole capital. Somebody could have seen this massive flow of capital if it really happened.
On the other hand unwinding the operation is trivially simple. Just use the same methods you use every 2 weeks to buy the coins and give them back to lenders and the massive debt and interest costs vanish instantly.