f-tilt u-tilt will work on the no tech and on any characters who have a fall speed slower than Lucario (24/41 characters). Those characters don't hit the ground out of f-tilt and therefore have no opportunity to tech: it's a true combo at all percents. Bowser's actually an anomaly here, but I think his weight might also be influencing that combo. The comprehensive list is on my KO% guide.
That's exactly what boost grab is. It's the same inputs as the DACUS without the upward Control Stick shift, basically. Experiment with dash grab and boost grab a bit side-by-side to see their comparative distances: boost grab has a bit more reach overall.
That's precisely why d-tilt is sometimes good. the percentages where u-throw sends an opponent too high for a follow-up are precisely the times you want to use d-tilt: so you can follow up. Example: Marth dies to tipper UAir on WarioWare at something like 97% out of a full hop (this is somewhere between the first tier of platforms, where he dies at 104%, and the top platform, where he dies at 93%). But at this percent, u-throw sends him well above the top platform. I'm not positive on this because I haven't actually tested it, but I'm pretty sure that d-tilt -> u-air will link on Marth at such a percent and allow for a KO. That's a really rough and simplistic example but I think it illustrates the point well enough. If not it'll definitely link into bair at other percents which can KO super early (~80%) on that stage.
Following up with fair is literally just a techchase, but you definitely need to Fastfall/L-cancel the fair in order to react. A lot goes into this in that you have to have paid attention to their tech options in other situations or just need knowledge of how good a characters' tech options are, i.e. Spacies can immediately shine out of a tech so tech in place is really good for them whereas Snake has a huge tech roll which is good for him to use. The fair is sort of a read in that sense, but it's at least educated in that you can be fairly certain of what they're going to do when they do it. Now as for covering what they do:
For when you don't get the spike hitbox (meaning they get sent more or less straight downward):
no tech: jab reset into literally whatever you want, c4, dair jab reset, nair, f-tilt into the best option (be it 2nd hit, u-tilt, grab*, c4**, or another techchase), d-tilt, crawl tilt, et cetera. The no tech is super free. It's also pretty common because it's not the easiest tech to get.
*this is a true combo on the floatier members of the cast, although I'm not positive where the cut-off is. It probably works on ROB and definitely works on Samus, Mewtwo, Zelda...
**Similar to the above, this is a true combo on really floaty members (I only know Samus, Puff, and Zelda for sure), but it can also punish missed tech on the f-tilt as well as punish tech in place with FH C4 detonate (which could be followed with an aerial at certain percents; ex. down air at low percents to set up a regrab or C4 stick).
Tech roll away: Tranq. It's just about the best option at this point because fair hurts and they'll be asleep long enough to at least get a c4 stick or another launcher hit. This is probably the second most common option because they tend to be DIing for the Uair, but you could also condition them to not do this by bairing them instead.
tech in place: regrab, tranq, c4, cypher... cover their fast options and time it right and you should still get just about anything. Pay attention to what option they tend to pick when they tech in place and use that against them: C4 beats shield, regrab beats most things, and cypher will at least trade with shine, and they'll probably not DI the cypher so if you don't get hit you get a C4 stick for free too.
Tech roll in: Tranq. Covers it even better than tech roll away. You can also probably DACUS.
No tech roll: tranq, DACUS.
get up attack: I have the most trouble covering this option honestly. I'd say shield it and then Cypher OOS as that's your fastest option and leads into most aerials or a C4.
This is all pretty much conjecture and thought experiments 'cause I haven't documented how I follow up on any of these (although I should) but I'm pretty sure this is what I do and it tends to work well.