I recently worked at one of the biggest MTB stage races in the world, the Cape to Cape in Western Australia. 1100 competitors over 4 days of fantastic trails and scenery.
I did the timing with RFID Ultras and a combination of ground mats and patch antennae. The ground mats are definitely the better option, in high-volume starts the side antennae tend to miss the occasional rider. So I use side antennae as spotters, and for one of the starts where ground mats were not possible.
Each rider had a tag on a seat-post sticker, and another on the front plate. I was hoping to do a comparison between the two tag positions, but the reality of race timing prevented that: competitors damaging tags, not putting them on correctly, losing them while riding, walking around the finish area near antennae, etc. For example, one of the riders had his seat-post sticker wrapped around the post, and only his front plate tag was read, and even then only by the ground mats. Racing environments are not at all suited to doing tests like that.

Anyway, without trying to sound like I'm bragging, the new RaceTec stage race functionality really made the scoring of this race a breeze. In past years I have set it up using dead zones, which worked well, but is not as elegant as a stage race setup. The final web results, as a stage race setup, can be viewed here. I spent very little of each day doing anything in RaceTec, just data checks and generally making sure things were running smoothly. The organisers used RaceTecWeb to get the stats they needed and for doing drop-outs, and the competitors had a kiosk where they could check their times (and compare over previous days). So I was left alone most of the time. I wish all race timing was like that!

Then this last weekend I worked at what is probably the biggest triathlon in the world in Noosa, Queensland, Australia. Over 5,000 individual competitors and nearly 3,000 team members (for the relay option), all racing on a single Sunday. I should use the word "at" in quotes, because I was actually sitting in my office, connected remotely to the race site managing data and systems, while another team worked on the ground. I used Teamviewer to remotely access a laptop in the timing office. And at the same time I was assisting another timing team working at an Ironman 70.3 in Taiwan.
It's quite difficult to work like this, you don't realise how important it is to have visual and auditory feedback at races. For example, has the commentator stopped calling names? Are spectators cheering because the leaders are about to come in from the bike leg? Is there someone standing at the door because there is a problem on course? I knew none of this, I was in total silence, sipping coffee, just watching RaceTec functions on a screen from 3,600 kilometres away. Very peaceful and stress-free, but a bit disconcerting. My client who contracted me to do this (and who was the one at the race site) is wanting to work like this on a more regular basis, and it will be interesting to see how it works out.