RTVega-neutral calendar in the index space is a complex bet. It assumes that you see value in term structure (i.e. fwd vol is cheap) and you see negative expectation in owning convexity (that is, gamma is rich). In that case, you can reasonably expect your vega leg to protect you from the convexity losses (somewhat) and have reasonably low expectation of convexity losses themselfs. Neither of these two conditions are true right now, if anything term structure is as flat as it can get in this low vol environment and gamma looks like a perfect buy to me. Just sayin.
95% of my calendars are with non-index underlyings. I constantly monitor the term structure of a number of underlyings. Eg. AAPL is set up for a calendar as I type which isn't helping my current position

