Allows the block size to increase exponentially to the point, where home-based nodes will no longer be able to participate in consensus.
I don't agree with that assessment, for four reasons:
1) Technology continues to improve. e.g. $100 worth of hard drive space 15 years ago costs about 17 cents today.
2) Block size cap != block size. The cap is 1MB now, yet we have no 1MB caps. The blocks will increase as needed to store the transactions that we have. Assumptions of 100% filled blocks are baseless.
3) If the increase based on transactions is still judged too quick, the miners can always choose to include only some of the transactions, based on size, priority, fee, their mood of the day, etc.
4) BIP 101 is not the end-all be-all. It's a best-guess shot in the dark. If it needs to be adjusted, it can be. It can even be adjusted down with a soft fork.
Hearn and Gavin made a mistake by releasing XT with BIP 101 with a 75% activation rule. They made a further mistake by including unrelated changes in XT. We should not let these issues confuse the examination of the block cap increase BIP alternatives. The simplest of the alternatives that actually has a chance to meet demand is BIP 101.