I like the brain-as-necktop meme. It is Dennett's way to dramatize the similarities between brains and computers. Desktop, laptop, palmtop, necktop. The brain-as-computer metaphor gets some criticism from philosophers - but the basic idea that the brain is functionally an information processing device, something that accepts sensory inputs and transforms them into motor outputs - seems simple and it ought to be fairly uncontroversial.
Memes being like apps seems a bit of a stickier analogy. I think it is fair enough to portray culture as being software for the brain. Not all brain software is culturally-transmitted (some is the product of individual learning). Also, some items of culture we might prefer to call data - rather than software. However, a broad interpretation of the term "software" can include data - so that seems like a minor nitpick. A more significant disanalogy involves complexity. Memes, many say, are simple, almost atomic bits of culture. Apps, generally speaking, are large and complex. There are other terms for a bunch of memes: memeplex and memome. Apps seem more like these than they are like memes. As with genes there's a bit of a philosophical quagmire over how big memes are. G. C. Williams once proposed that genes needed to have an 'appreciable frequency' to qualify - the idea being that this rules out entire genomes - since they are unique in the population and therefore are typically not very "frequent". Apps actually pass this test - since the high-fidelity copying found on the internet means that apps are often identical to other copies of them - down to the last bit. So: apps have a meaningful frequency in the population of all apps. However, this seems more like a limitation of Williams' criterion than a legitimate reason for identifying memes with apps.
Memes being like apps is OK - in that both are types of software. Perhaps it's an analogy that shouldn't be pushed too far, though. It might be better just to say that apps are made of memes.