Bitcoin is slave to fiat for as long as the merchants predominantly want fiat, because this drains fiat from the ecosystem as investors bring fiat into the ecosystem.
It's not only "investors" that are bringing money in the system but also any person who is converting Fiat for bitcoin with the intent to use it to make payments. In this case it is closer to a currency conversion. If I sell my EUR to buy USD to buy something in the USA, I am not "investing" in the USD.
If people are happy with the way the transaction happened, they are likely to convert their fiat into more bitcoin for later use. Probably a bit more than they used last time because past experience increases confidence in the technology and perceived utility.
About the investors and their use of bitcoin, of course they are less likely to buy-back right away after a purchase at a time when the price is going down. For this reason the point of Risto makes sense in that it only makes the reversal more pronounced when all the people who spent coins feel that we have bottomed, price keeps rising and they have to panic buy.
If Bitpay grows 3 times faster than adoption, the question lies in the rate of replacement amongst the buyers with bitcoin. Do you know how the bitcoin position of consumers in bitcoin is evolving ? What if they would be re-converting as much as they spend ? Then even a stagnant adoption would mean a stagnant price, not an endless drop. And a growing adoption would mean a rising price. And what if existing bitcoin consumers are actually converting more fiat than last time at each conversion ?
The question is : do you have data that proves that the bitcoin going through bitpay are "getting rid of", as opposed to "being used and replaced" ?