They are saying however, that the monero blockchain is bloated and that transacting could become very slow in the future because of the large database of transaction information.
Sure, it's an absolutely valid concern.
I think the important thing is not to compare Monero's block size with Dash, but firstly with Bitcoin so that we have a baseline. Monero's transactions are smaller than Bitcoin's, like-for-like. So if Monero were a "transparent blockchain" like Bitcoin's we would have a
smaller blockchain than Bitcoin for the same transactions! But since part of Monero's "opaqueness" involves hiding the origin of transactions by making inputs appear to come from multiple transactions, this incurs an additional size penalty. This means that, for the same number of transactions, Monero's blockchain is about 4-5 times bigger than Bitcoin's.
It is important to note, though, that any privacy feature will have a similar impact. Dash's DarkSend, centralised Bitcoin mixing services, or even decentralised initiatives like JoinMarket (which is built on the CoinJoin scheme) all have an increased transaction size. If they are enforced on every transaction (ie. on a protocol level) it means the entire blockchain is significantly larger - and that applies to Bitcoin too!
The reality when it comes to blockchain sizes is that
linear blockchain bloat is largely insignificant. What I mean when I say that is this: which is more convenient to host, a 350gb blockchain, or a 1.58tb blockchain? Well they're both inconveniently large, and whilst you could argue that a 350gb blockchain fits on a 512gb SSD today (where a 1.58gb blockchain doesn't) it's obviously growing so fast it'll outstrip that storage space in a matter of months. Basically any blockchain above, like, 20gb is a complete pain in the posterior.
The solution to this problem is
not to try incremental, linear decreases in blockchain size. If you have a useful cryptocurrency that people are using
the blockchain is going to be large. Suck it up, embrace it, accept it right now. Rather, the long-term solution can be found in blockchain pruning, whereby we cut off all the historical data and only keep a subset (the utxoset for Bitcoin, and the txoset + key image set for Monero). In this Bitcoin has a slightly different approach to Monero, where nodes are either "archival" (keeping the entire blockchain) or "current" (keeping a small portion). With Monero we will be implementing pruning in a per-node configurable basis, where node operators can choose to either keep the whole chain (default), keep only the last M blocks, or keep everything from block N. This will mean that, as a node syncs up, it will find progressively more nodes that have the blocks it is requesting, until it reaches the top block in the network, and every node has that available:)