it may even have been good. except that it is encouraging this non-verification scheme for tx's which as you say, may be gamed and has contributed to quite a perversion in analyzing this particular attack and was never visualized in Satoshi's original ideas.
That isn't the problem.
The problem is that there is no way to tell an SPV client that the chain they are following because it has the most proof of work is actually invalid and should be rejected.
If that capability existed, then nobody would have to care whether or not miners choose to burn their own electricity mining invalid blocks or not.
This seems to happen frequently in Bitcoin where bad behaviour by party A can negatively effect party B, and so everybody focuses exclusively on preventing party A's bad behaviour instead of making the system more robust by removing party A's ability to negatively impact party B to solve all current and future problems.