1. Profit switching should also depend on the profit change % (e.g. Nicehash uses %10 to switch to another algo). If any drops happen over this value it should stop and change the miner.
   2. While stratum connection fails for some reason for a long time it should not try infinitely to connect (sometimes you get banned or server could be down), it should switch to the second or third best alternative after some tries (e.g. 5, 10 depending the user input)
   3. For non-homogeneous systems (e.g. with different gpu's with different settings), card by card benchmarking should be available (like nicehash). And depending on the best profit result, every card or card group could mine the best profitable one on itself (e.g. nicehash redirects my 3 asus dual's to dual mining, 2 msi's to equihash, 3 evga's to .... )
   4. When I try to assign the hash rate value coming from the mining process (incremental value in a specific time), it overrides the single card value and writes over the sum of the hash rate onto the current single card's benchmark value. I do not know if it changes the profit calculation but it should be corrected.
   5. Algo/gpu dependent benchmarking could be available. As in nicehash example, it benchmarks every miner with every gpu. It tries both excavator and ewbf for zcash mining; claymore and ethminer for ethereum mining; alexis and tpruvot for keccak etc. and selects the most profitable one while mining. It is done at first while benchmarking.
   6. There is some problem with the alexis version of ccminer. I tried both manualy and with awesome miner but it gives error with my gpus.
   7. Switching to ewbf is not easy, it is getting opened and closed too many times in a short period of time in order to start mining. When I try manually from the bat file it goes perfect.
   8. In longer times, second or third instances of ccminer could start. Both of them mines while lowering the performance of themselves. Third instance is generally passive.