With an average monitor what you see on-screen is already soft proofed to sRGB (or something very close to it), because that's all the monitor is capable of displaying anyway. So soft proofing to sRGB won't tell you anything. You won't see any difference.
In Photoshop it sounds as if you assign profiles. That's not the way to do it. If you convert correctly you won't see any difference. Same principle as above: there may be clipping in the process, but what you see on screen is already clipped, so no visual on-screen difference.
With a wide gamut monitor soft proofing becomes slightly more useful. But still you won't see any changes occurring outside Adobe RGB. You'll get a better idea by keeping an eye on the histogram. Ideally, all three channels should taper gently off towards the endpoints. If any one or two channels are backed solidly up against the endpoint, on either side, that's gamut clipping.
If Blurb gave you a real profile, one that reflected their actual printing process, you could soft proof to that. But apparently they don't.
As 21 has explained, based on the Blurb "workflow" it is pointless to try to soft proof! They don't provide the profile used to convert the data.