Well, there are strategies which certainly are more dependent on dice rolls than others and games can therefore be decided by rolls (imagine an "all in" style like e.g. Rushmaster, rolling blanks on two boulders). But that is part of the strategy and you have to be aware of the probability to fail with this kind of play.
But some points the reddit OP mentions i do share (and i played a lot of MW games). If you play for fun in the sense that you play thematic books, without putting too much thoughts into your books and try to optimize - briefly speaking without any competitive ambitions - MW is perfectly fine. Now, if you try build maximized books you'll find a huge number of cards every mage absolutely has to have. For some cards, there are no alternatives, which makes some mages intrinsically inferior compared to others (because the pay more for these cards or have less of them) which can be a source of frustration. Did you ever try to make a cool strategy on a lesser played mage work and then recognized that the other mage is just better in almost every aspect? Happens to me a lot.
It's leading to a small pool of mages being played and a small variety of cards being used and finally in a small variety of strategies. For example, I can safely say that over 90% of our games are finished without more than one or two creatures being killed, since focusing the mage is just more efficient. I think the game would profit - and that is a point the reddit OP also states - from more mage exclusive solutions to common threads/problems (e.g. getting rid of armor, dispelling/countering enchantments, gain/deny mobility).
Therefore, I can understand some points mentioned in the reddit post, but I'm optimistic - I think some points exist because the card pool is simply not big enough yet.