I've worked with Cisco UCS blade servers quite a bit and have no idea why there would be any difference in security requirements than rack servers. They're both just compute with an out of band management and utilize the same network principles.
The main advantage of blade servers is that you can:
- Aggregate compute into a smaller formfactor
- Multiplex their connectivity through the chassis backplane for (in theory) faster east-west communication
- Have less individual connections to a TOR or DC switch.
- Typically, they have a "profile" type of management software that allows for quick and easy replacement of blades in the event of a failure.
In my opinion, these are the considerations you need to take into account, especially if connecting directly into a Meraki switch. What I often see people get wrong is underestimating (or not fully understanding) their network requirements for the blade servers and the oversubscription on the physical links (especially in a converged network)