With the recent release of Dark Magician support, it's again confirmed disagreements I've had with Konami's card design philosophy for a long while.  I like to think that Konami can make high profits and balanced play at the same time by improving their integrity and broadening the base of players, but that might be wishful thinking.

Hype Waves

The first issue is that support for an archetype is generally non-existent for 5-10 years and then it is released in a giant wave.  This appeals more so to hype than sound game design: make fans hungry with a drought and then excited with a big wave of support.  The deck becomes an overplayed trend for a month or two, and then it is largely neglected.  Once the archetype is made obsolete via power creep, it receives nominal support every so often, but this support is usually inferior to the older support and rarely addresses the archetype's flaws to begin with.

This is problematic for people who have ran the deck for years when it was low-tier ; everyone hops on the train and that person who originally ran it is no longer unique.  Not only that, but now several people at that person's locals is running that deck.  While the person was originally able to win locals with their rogue build, now they must alter their deck to conform to the generic cookie cutter standards just to remain competitive.

Another byproduct of this is that we often get a few overpowered support cards at once instead of a wide variety of balanced support released over time.  If Plants had steady support, Lonefire Blossom would have had limitations.  Substitoad would've had a once per turn clause if Frogs were a viable archetype before it came out.

Lack of Variety

The second issue is that archetype support is made so that every player should run the deck essentially the same.  Deck building skill is minimal and the only real choice you have is the ~5 deck slots left after the essential cards and overall staples are used.

Archetypes should have more than one way of being ran.  It would certainly be a step up to allow an either-or scenario.  Offensive or defensive build?  Aggro or control build?  But even more can be done with that.  Allow every step of the way to have an element of choice.  One good method would be to make two cards with a similar function, but treat them both as the same name (i.e. Cyber Harpie Lady vs. Harpie Lady 1).  

With Dark Magical Circle, for example.  They could release "Dark Magical Sphere", which is treated as "Dark Magical Circle", but it has a much stronger search effect in exchange for a removal effect which merely destroys instead of banishes.  In this case, you could do 3 Sphere, 3 Circle, 2 Sphere/1 Circle, 2 Circle/1 Sphere.  Do this with several different cards and players can eventually find an intricate balance which promotes a specific kind of play.

The worst case scenario, with many archetypes, is that the Spell line-up (for example) consists of 4 staples, and 3 playsets of obvious archetype support.  There's no tradeoffs.  There's no complexity.  There's no individuality or choice.