Optimize in-flight operation cleanup method

The current approach to ensure in-flight operations aren't left dangling is to create a Timer for each operation which then marks the operation as canceled. This is somewhat redundant, given that each operation already has a much shorter operation timeout. Instead, we can implement a more lightweight approach to ensuring we don't leak memory for in-flight operations.