The problem with running numbers like this is that it ignores opportunity cost. Let's say you're basing this on a $5.95 cost per $500 gc, earning 3000 HH for a cost of just under .2c per point. At this cost, a 40k redemption is about $80 in real cost (~14 gcs). Alternatively, you could have earned 5% back on a different card and earned $350. So for the same cost, though you can manufacture those 40K HH points at .02c per point, you could have had $350. For those HH points to not be a losing proposition when compared against that, you'd need to be earning .8-.9 cents per HH point, which is difficult to achieve.
The SPG card makes even less sense in this scenario to use at grocery stores, but you wouldn't use it there. You'd likely try to MS with SPG at a non-bonused cost of $2.95 per $500, at where you could net, say, 12k at a cost of $70 (cheaper than your grocery scenario). Even funnier, you could MS with HH at this same non-bonused cost of $2.95 per $500, only earning 3 points per dollar, but actually costing less to earn those 40kHH (at just under $80).
What your argument presumes is that there is lost opportunity cost. I max out my 5% opportunities first, then my mall spend, then do my other MS. I'd never forgo 5% cb for any other MS (I hit 5% as hard as I dare).
HH at 6x grocery is indeed higher cost than at mall (as I indicated), but there's the issue of volume. My grocery volume is effectively unlimited and I get 6x HH, while mall @ 3x HH is limited. If I did HH at mall, I'd need to do 2x the volume and, even worse, I'd crowd out my ability to earn low cost SPG. I accept modestly higher cost HH at grocery in exchange for otherwise-limited ability to earn low cost SPG elsewhere. And since I value SPG for it's optionality, I consider it the best solution and a net gain.
As such, I don't run my cost analysis based on artificially low cpp for SPG, and instead use a blended cost so as to not skew the results (and I assume higher cost cpp for HH). The upside is I have access to meaningful number of SPG, allowing for stays at properties where HH isn't present, or aren't as desirable. How do you place value on that?
Finally, there's the ability to earn status. I don't spend a lot of nights in chain hotels (~20 nights a year). Not enough to get top status via stays. Yet HH offers Diamond for $40K spend on Surpass (and Reserve). No other hotel program does that. I've had excellent upgrades as HH Diamond (better than my experience as Hyatt Diamond). With low cost MS and ability to get top status, coupled with large property list, HH is almost always in contention for my business. But if I find better properties/preferred locations (even when at higher cost), I will stay elsewhere.
But my point wasn't to get in a debate whether SPG or HH was lower cost. It was that everyone needs to run their own numbers to see what is best value for them based on their own spend patterns, and not listen to emotional comments that infest this hobby every time there's a devaluation. Reality was HH was so obscenely generous prior to the devaluation that it wasn't that there was such a large devaluation, but rather it took so long for the inevitable to happen. Pre-deval, you could get top tier rooms for $50/night - rooms with rack rates btw $600-$1K/night. Now, it's closer to $150-180/night (via 6x grocery and 5th night free) - a "horrible" percentage difference in the abstract, but compared to the competition (Hyatt, SPG/IH etc), remains very competitive value.