That's a very good question. I think there are at least three different aspects to it: philosophical, psychological and practical.
The first category is one which I am more than happy to leave for others to ponder.
Psychologically, I have witnessed vastly different attitudes toward other living creatures: from people treating pets as full-fledged family members, through indifference to outright cruelty. The kinder, gentler attitude seems to be a modern phenomenon, more common the further removed one is from the realities of the food chain (a "fun" factoid I just looked up: some 25 million chickens are slaughtered every day in the US alone, and I think we all know how most of them live; there is nothing kind or gentle about it). There is a widespread belief that kids benefit from growing up with pets, and some evidence that it may actually be true. Based on what I saw growing up with a sequence of dogs, a bunch of birds, miscellaneous rodents and the occasional reptile, I think it would be pretty straightforward (and more humane - I was a lively kid...) to replicate the experience using virtual pets. It may be hard to fool a real dog into thinking that a simulated one is real, but fooling a human is pretty easy.
The practical aspect is the serious one. Flavio Mercati has written an essay which pays homage to currently fashionable views (and which therefore can be expected to do very well in the contest) and advocates "solutions" like only eating game meat, but which also gets some things right. One of them is that there are vast amounts of energy of raw materials awaiting exploitation in space. Another is that the biological diversity of Earth is truly rare, and therefore precious. He is sticking to the 60s script of space colonization, with humans turning other planets into new Earths by terraforming, and in that model, the more species you have to work with, the more likely you are to find a viable mix capable of supporting a robust biosphere in the new environment. The script is dated, but he does have a point: diversity is good.
That won't matter to inorganic Humans 3.0, but it could bite Humans 2.0. Their life support system will need to provide them with things like glucose and amino acids. Ideally, those would be synthesized, but initially at least they could come from a handful of plant species (Soylent is essentially based on rice, oat, canola and microalgae; in The Millennial Project, Savage was big on blue-green algae). Since they will be working with a completely engineered environment, it will be tempting to optimize everything, down to cloning a few particularly productive organisms, and call it a day. The result would be a monoculture, with all that entails: very efficient as long as it works, but very fragile when something goes wrong. So as a purely practical matter, it would be in their interest to maintain a larger selection of species.