Is your argument that $1500 / mo is too much? Why would the engineering team not be more rigorous in their model selection given a constraint?
$1,500/mo * 14 months = $21,000.
If local models are 14mo behind as many in HN say it may be profitable to just wait. Maybe just spend a few hundred dollars of your tokens and buy hardware piece by piece.