I am eager to see a followup on this after 10 months.

In the mean time, ETH price reached an ATH of $1380 from $230 and dropped back to $400 as of writing. Whereas, ETH mining difficulty increased 468% in the past 9 months. I always believe, Cloud mining is lesser profit than buying the coins directly. Because, people who promote cloud mining always undermine the difficulty increase. But the reality is, if a coin is very profitable, the difficulty level increases parabolically in the next few months and the profitability drops severely.

In this case, the author of the article mentioned that due to difficulty, he is going to get a bit lesser than projected ETH. However, In today's difficulty, it will take 1 year to generate 1ETH using 37Mhz Mining power.

Using a very conservative difficulty increase (9% per month), he is going to get less than 6 ETHs in 2 years.
Using the actual difficulty increase(468% per year/12 = 39% per month), he is going to get less than 2 ETH in 2 years.

This is the ugly part every Cloud mining promotor misses. Difficulty increases 100s% and makes it far less profitable than buy and Hold. In this case, the user will end up getting less ETH in small pieces instead of having more ETH and an option to sell it any time.
In very rare occasions, cloud mining gives more profits.

My Conclusion: Just like last year, this year also is better for buying and selling. Could mining was/is/will not (be) profitable than buying and selling directly.