I wouldn't expect you to agree. I've seen this argument played out elsewhere and I don't imagine my feelings on it will much matter to you or anyone else here. From what I can tell, it's largely a "because I can" effort, so I won't get into a point-by-point argument of where we disagree.
Thankfully, most open source developers, and most open source-oriented sites, take a completely different approach.
Why? If I do so, the site is not able to count how many distinct users have downloaded the class.
It's a meaningless statistic anyway because a download can't be equated to a "user". Personally, I probably download 15-20 different classes/procedures/samples for every one that I actually ever use in a project. If I do use it, I let the author know. I suspect that most other developers are the same way. I know there are people out there who build entire sites with nothing but downloaded code, but I don't consider that software development.
Further, there are lots of ways to measure downloads without requiring everyone to register before they download.
Subscribing does not cost anything.
No, but it's a site that wants my email address for a resource that I can typically find elsewhere without having to supply give up any contact info.