I use the Class factor 213 (or 13 rClass) in my contender selection process. To separate contenders, I don't look at up or down, rather I rank the horses by class and look for a large gap/break. I'll give it a try explaining.
Percentages & Probabilities 2022, Dave refers to this as the "King of Class" (about the 8:26 mark of the video). It's comprised of factors 14, 15 and 16. 14 - Earnings per Start (based on a % compared to today's purse). 15 - Earnings Box Consistency (6 points win, 3 points 2nd, 1 point 3rd divided by the number of starts times 6). 16 - Average Purse Value (based on a percentage compared to today's purse). I think I got this right based on the video.
I like to use the Class factor to separate contenders especially if there's a large gap. That's what I look for. Examples:
Today 08JAN2025 TAM Race 4. Sort and rank by class-- the top ranked horses are 65, 62, 53 and then there's a gap/break to the 4th ranked horse at 34 (19 point gap). To me these top 3 horses have the best class. The #1 Mamba with the 65 class rating won and paid $8.00.
Today 08JAN2025 TAM Race 6. Sort and rank by class-- the top ranked horse is 60 and then there's a gap/break to the 2nd ranked horse at 44 (16 point gap). The #3 Like with the 60 class rating won and paid $14.20.
I like to see large 15+ point gaps/breaks. These horses don't always win, but if they qualify under other factors like Pace, PSR and Trainer-- I have what I believe is a solid contender.

BRIS class rating

Look at the new 25,000 series of factors.

https://www.community.horsestreet.com/discussion/comment/7266

This is a good question as there are many forms of class to consider: purse, claiming tag, track (is the horse shipping in from a higher or lower class track?), trainer, or jockey. These considerations aren’t based on time but are certainly reflected in the odds. My question is when are these factors more predictive than past performance data?

I'm not sure that any of these factors can be used by themselves and be more predictive then the key past performance data as a whole. While a contender may have an advantage in 1 or 2 key factors, they may rank lower in others. I think you have to look at for example-- speed and pace with these other forms of class.

We have many options available:
Class level of last good race
class level of best good race
Average competitive level
Back class
Earnings per start
Earnings box class
Composite earnings-based class
Composite level-based class
Time-decay class performance
Time-decay average purse value

And, like you said - is speed just another way of quantifying
class? An the others also.

We have more than enough ways to determine it. Is it just a
personal choice on which one you use? Which do you think
is best?

Which is the most profitable? Need to do some research on that.

I think my post might’ve veered off the subject a bit. To answer the original poster’s question: in HSH, I use rcClass, but I’ve never used class to select contenders. I’ve used it to break ties among contenders or add a longshot contender if the horse is ranked first or second in rcClass.