Christian, author at StellarX, compares this property to the Great Filter. Many ICOs had failed to realize what Ethereum is really good for, which consequently led them to an eventual shutdown.
Earlier this year, the StellarX team conducted a research. They used a test app to generate a specified number of transactions from numerous accounts. The app ran for a total of 10 hours during the periods when gas prices were cheap. As a result, they spent $13,000 on transactions fees.
Christian says that Ethereum is a great choice for ownerless applications that need a decentralized decision-making mechanism. Augur is a perfect example because prediction markets largely benefit from an automated way of establishing the truth about real-world events.
However, Ethereum is not the go-to choice for those applications that need to handle a large amount of transactions between users. The team set the test app to generate 1 transaction per minute from 480 accounts for 3 hours, which totals to 86,400 transactions with an average of 8 per second. When the test was over, more than half of the transactions were still missing. The team examined the results and discovered the fact that the more transactions an account submits, the longer it takes to confirm them.
Further research found out that individual queues are not the only issue with Ethereum. The research team proved that Ethereum cannot serve as a platform for those applications with a large passive audience, where users send a single transaction once in a while. As their next test, they set the app to submit a single transaction from 28,800 accounts. As soon as the transaction appeared on the network, miners’ fee went up. The earlier transactions were too cheap for miners to include them and they had to wait for hours before getting confirmed.
As a conclusion, Christian states: